Prediction model for dynamically predicting treatment effect of Peg-IFN-alpha combined with NAs and application thereof

Abstract

The invention relates to the field of medical detection, in particular to a prediction model for dynamically predicting the treatment effect of Peg-IFN-alpha combined with NAs and application of the prediction model. According to the invention, different key predictive factors are screened for CHB patients in different treatment cycles for the first time, that is, a model for dynamically predicting the curative effect of the CHB patients is established on the basis of various hepatitis B curative effect index data obtained at different time points when the CHB patients receive Peg-IFN-alpha combined NAs treatment, so that the clinical outcome is dynamically predicted along with the treatment process of the patients; the method has good prediction efficiency, and is of great significance in reducing economic burden of patients, reducing adverse reactions as much as possible and optimizing treatment strategies.

Description

Technical Field

[0001] The present invention relates to the field of medical detection, and in particular, to a dynamic prediction model for the therapeutic efficacy of Peg-IFN-α combined with NAs and its application. Background Art

[0002] Chronic hepatitis B (CHB) usually cannot achieve complete cure by completely eliminating the hepatitis B virus. The current goal of clinical treatment is to achieve clinical cure (also known as "functional cure") by maintaining negative hepatitis B surface antigen (HBsAg) after stopping treatment. The current drugs for treating CHB include nucleoside analogues (NAs) and interferon (IFNα). NAs alone often cannot achieve the clinical cure goal of HBsAg negativity, but NAs-treated CHB patients combined with peginterferon α (Peg-IFN-α) can partially achieve clinical cure, which is one of the important strategies for achieving clinical cure of CHB. When NAs combined with Peg-IFN-α are used to treat CHB patients who have been treated, achieving continuous negative HBsAg is the key goal to achieve clinical cure.

[0003] In order to achieve clinical cure, CHB patients often need to receive long-term drug treatment, and a high proportion of patients still cannot achieve clinical cure after long-term treatment. Taking the combined treatment of NAs and Peg-IFN-α in CHB patients with good treatment as an example, patients usually need to receive continuous drug treatment for 2-3 years to achieve HBsAg negative conversion. Currently, only no more than 30% of patients can achieve clinical cure with the best NAs and Peg-IFN-α combination regimen. In addition, long-term use of Peg-IFN-α is relatively expensive, and Peg-IFN-α is prone to adverse reactions (such as influenza-like syndrome, bone marrow suppression, mental disorders, autoimmune diseases, retinopathy, kidney damage, etc.). Therefore, when the goal is to achieve clinical cure of CHB, it is of great significance to predict whether CHB patients can eventually achieve clinical cure based on the early treatment response level, which is of great significance for reducing the economic burden of patients, minimizing the occurrence of adverse reactions and optimizing treatment strategies.

[0004] In recent years, people have been committed to exploring methods to predict whether patients can eventually achieve HBsAg negativity in the early stage of CHB treatment. Previous CHB studies have shown that the absolute value of HBsAg after treatment and the change relative to baseline are key predictors of HBsAg negativity. A study predicting HBsAg negativity in 1,240 CHB patients [1] showed that the HBsAg baseline value and the rate of HBsAg decline within 48 weeks were key predictors. However, most methods used to predict HBsAg negativity in CHB patients are based on research data on NAs monotherapy or combined with multiple treatments, and no prediction method specifically for Peg-IFN-α combined with NAs to achieve clinical cure in CHB patients has been reported. In addition, existing prediction models mostly predict endpoints based on observations at specific time points, and do not establish a prediction model that matches the treatment cycle according to the treatment process to screen different key predictors for patients in different treatment cycles. Summary of the invention

[0005] In order to fill the gap in the prior art, the present invention aims to provide a dynamic prediction model based on commonly used clinical hepatitis B examination indicators for dynamically predicting the clinical efficacy of Peg-IFN-α combined with NAs in the treatment of CHB patients.

[0006] Specifically, the present invention will intercept data of different treatment cycle lengths to establish a series of prediction models to achieve dynamic prediction of clinical outcomes as the patient's treatment progresses.

[0007] The first aspect of the present invention provides a group of biomarkers for dynamically predicting the therapeutic effect of Peg-IFN-α combined with NAs, wherein the biomarkers include Height, V_HBsAg 8W , P_HBsAg 8W , P_HBsAg 36W 、P_HBsAb 36W , D_HBsAb 48W , P_HBsAg 48W 、P_HBsAb 48W , P_HBsAg 60W 、P_HBsAb 60W 、V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W .

[0008] The second aspect of the present invention provides an evaluation model for dynamically predicting the therapeutic efficacy of Peg-IFN-α combined with NAs, wherein the model uses the aforementioned biomarkers as key evaluation indicators.

[0009] In one embodiment, the model evaluates key indicators of the patient's samples according to the different treatment cycles of the patient, so as to predict the therapeutic efficacy of the patient. The treatment is 12 weeks as a treatment cycle. In one treatment cycle, Pegibine is treated for 8 weeks and then stopped for 4 weeks. NAs are continuously used for a total of 12 treatment cycles, wherein:

[0010] The calculation formula for the prediction score of the first cycle is:

[0011] L1=13.124-0.083×Height-0.002×V_HBsAg 8W -0.673×P_HBsAg 8W ;

[0012] The prediction score calculation formula for the first two cycles is:

[0013] L2=12.915-0.081×Height-0.002×V_HBsAg 8W -0.672×P_HBsAg 8W ;

[0014] The prediction score calculation formula for the first three cycles is:

[0015] L3=-2.873-0.539×P_HBsAg 36W +0.648×P_HBsAb 36W ;

[0016] The prediction score calculation formula for the first 4 cycles is:

[0017] L4=-3.651-0.044×D_HBsAb 48W -0.739×P_HBsAg 48W -0.685×P_HBsAb 48W ;

[0018] The prediction score calculation formula for the first 5 cycles is:

[0019] L5=-3.862-0.616×P_HBsAg 60W +0.416×P_HBsAb 60W ;

[0020] The prediction score calculation formula for the first 6 cycles is:

[0021] L6=-5.113+0.008×V_HBsAb 60W -0.811×P_HBsAg 72W +0.028×P_HBsAb 72W ;

[0022] The evaluation model is used as follows: when the patient is in the nth treatment cycle (1<n≤6), the L calculated based on the corresponding key indicator data n The prediction score of L n The critical value is used to determine whether the patient has HBsAg negative conversion at the end of treatment.

[0023] In one embodiment, the corresponding critical values ​​of L1, L2, L3, L4, L5 and L6 are 0.3465, 0.3474, 0.2340, 0.2269, 0.4166 and 0.1936, respectively. If the value calculated by the model is higher than the corresponding critical value of the cycle, it is predicted that the patient will eventually be HBsAg positive (HBsAg has not turned negative), otherwise it is predicted that the patient will eventually be HBsAg negative (HBsAg has turned negative).

[0024] The third aspect of the present invention provides an application of the aforementioned biomarker in the preparation of a product for dynamically predicting the therapeutic efficacy of Peg-IFN-α combined with NAs, the product comprising a kit, a chip, a nucleic acid membrane strip, a diagnostic device and a diagnostic instrument. Preferably, the product is a diagnostic device and a diagnostic instrument.

[0025] A fourth aspect of the present invention provides a diagnostic system for dynamically predicting the therapeutic efficacy of Peg-IFN-α combined with NAs, the diagnostic system comprising:

[0026] (1) Detection module: The detection module is used to detect the values ​​of multiple biomarkers, wherein the biomarkers are selected from Height, V_HBsAg 8W , P_HBsAg 8W , P_HBsAg 36W 、P_HBsAb 36W , D_HBsAb 48W , P_HBsAg 48W 、P_HBsAb 48W , P_HBsAg 60W 、P_HBsAb 60W 、V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W The value of

[0027] (2) Result judgment module: The result judgment module is used to determine the biomarkers Height, V_HBsAg detected by the detection module. 8W , P_HBsAg 8W , P_HBsAg 36W 、P_HBsAb 36W , D_HBsAb 48W, P_HBsAg 48W 、P_HBsAb 48W , P_HBsAg 60W 、P_HBsAb 60W 、V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W The numerical value is obtained and the above-mentioned diagnostic model is used to output whether the subject is ultimately HBsAg positive or negative.

[0028] In one embodiment, the working process of the user submission module includes the following:

[0029] (1) The user fills in a form on the WeChat applet or web page, where the content is selected from Height, V_HBsAg 8W , P_HBsAg 8W , P_HBsAg 36W 、P_HBsAb 36W , D_HBsAb 48W , P_HBsAg 48W 、P_HBsAb 48W , P_HBsAg 60W 、P_HBsAb 60W 、V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W ; Fill in the indicator values ​​and patient baseline values ​​required for the score calculation formula for each cycle according to the patient's different treatment cycles:

[0030] (2) The front end (mini program or web page) obtains the form data filled out by the user and sends the data to the back end program through a server request. The back end program obtains the user's diagnostic data by receiving HTTP requests and passes it into the diagnostic algorithm.

[0031] According to a fifth aspect of the present invention, there is provided an electronic device, the electronic device comprising:

[0032] (1) Input unit: The input unit is used to input multiple physiological index data of the subject, and the physiological index data is selected from Height, V_HBsAg 8W , P_HBsAg 8W , P_HBsAg 36W 、P_HBsAb 36W , D_HBsAb 48W , P_HBsAg 48W 、P_HBsAb 48W , P_HBsAg 60W 、P_HBsAb 60W、V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W ;

[0033] (2) Analysis unit: The analysis unit is used to analyze and calculate the physiological index data input by the input unit according to the aforementioned diagnostic model, and obtain the prediction score of the subject, and further predict whether the subject will have HBsAg negative conversion at the end of treatment according to the aforementioned critical value;

[0034] (3) Output unit: Outputs the results predicted by the analysis unit through screen display, tactile feedback, color feedback, etc.

[0035] In one embodiment, the subject's physiological indicator data to be entered is selected based on the subject's treatment cycle.

[0036] A sixth aspect of the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program is a computer program for implementing the functions of the aforementioned evaluation model or the aforementioned diagnostic system.

[0037] Compared with the prior art, the present invention has achieved the following significant progress:

[0038] (1) NAs combined with Peg-IFN-α for the treatment of CHB patients is currently an important strategy for achieving clinical cure of CHB. In order to achieve clinical cure of chronic hepatitis B, CHB patients often need to receive long-term drug treatment, and a high proportion of patients still cannot achieve clinical cure after long-term treatment. At the same time, long-term use of Peg-IFN-α is relatively expensive, and Peg-IFN-α is prone to adverse reactions. Therefore, predicting whether CHB patients can eventually achieve clinical cure based on the early treatment response level is of great significance for reducing the economic burden of patients, minimizing the occurrence of adverse reactions and optimizing treatment strategies;

[0039] (2) Existing prediction models mostly predict endpoints based on observations at specific time points, and do not establish prediction models that match treatment cycles according to the treatment process. The present invention screens different key prediction factors for patients in different treatment cycles to achieve dynamic prediction of clinical outcomes as the patient's treatment progresses. That is, based on a variety of hepatitis B efficacy indicator data obtained at different time points in CHB patients receiving Peg-IFN-α combined with NAs treatment, a dynamic prediction model for the efficacy of CHB patients is established and its prediction performance is analyzed. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0041] Figure 1 ROC curve obtained for the modeled dataset;

[0042] Figure 2 Calibrability curves obtained for the modeled data set;

[0043] Figure 3 ROC curve obtained for the validation dataset;

[0044] Figure 4 This is a schematic diagram of the prediction applet. DETAILED DESCRIPTION

[0045] Here, the clinical trial data of the applicant's own product, long-acting interferon pegibrin combined with NAs for the treatment of CHB patients (clinical trial registration number: NCT04846491) is used to specifically demonstrate the implementation of the present invention. The present invention can also be used to establish a corresponding prediction model based on data from other CHB treatment regimens, and more clinical data of pegibrin combined with NAs for the treatment of CHB patients can be incorporated to further update the prediction model.

[0046] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0047] Example 1 Screening of predictive factors and establishment of prediction model

[0048] (1) Data collection on patients receiving combined treatment with peglibin and NAs

[0049] Data of 244 CHB patients who received combined treatment with pegibectin and NAs were collected dynamically. Serum samples of the 244 patients were tested before treatment, with 0.05IU / mL≤HBsAg<1500IU / mL, HBV DNA<100IU / ml and HBeAg<10COI. In the example clinical trial, the aforementioned patients received treatment in a 12-week cycle. In one treatment cycle, pegibectin was treated for 8 weeks and then stopped for 4 weeks. NAs were continuously used for a total of 12 treatment cycles. The collected information included demographic information (sex, age, height, weight), the time course of common hepatitis B venous blood test indicators (HBsAg, HBsAb, HBeAg, HBeAb, HBV DNA and ALT) and pegibectin dose. The collection time nodes of common hepatitis B test indicators were weeks 0, 4, 8, 12, 24, 36, 48, 60, and 72. According to the serum HBsAg level of patients after 12 cycles (144 weeks) of combined treatment with pegbin and NAs, patients were divided into HBsAg-negative group (HBsAg negative conversion, HBsAg < 0.05 IU / mL) and positive group (HBsAg not negative conversion, HBsAg ≥ 0.05 IU / mL). 2 / 3 of the data were randomly selected for the establishment of the prediction model (modeling data set), and the remaining 1 / 3 of the unmodeled data was used as external validation (validation data set).

[0050] (2) Univariate logistic regression to screen the time points of time-varying indicators

[0051] Some common hepatitis B monitoring indicators show trend changes after receiving pegbin combined with NAs treatment, which has a certain time process. However, time series with strong correlation cannot be included in the prediction model at the same time. Univariate logistic regression is used to screen the time point with the strongest performance in predicting the endpoint among time-varying indicators, and the Akaike information criterion (AIC) value is used as the judgment standard. After preliminary data analysis, HBsAg, HBsAb and ALT, the common hepatitis B indicators, changed significantly after treatment. The above three indicators and their derived variables (changes relative to baseline and logarithm of the ratio relative to baseline) were screened. Taking the HBsAg indicator at week 8 as an example, the indicators to be screened include the absolute value of HBsAg level at week 8, V_HBsAg 8W , relative baseline change value D_HBsAg 8W =V_HBsAg 8W -V_HBsAg 0W The logarithm of the ratio of the relative baseline value P_HBsAg = ln(V_HBsAg 8W / V_HBsAg 0W ). After grouped univariate logistic regression (Table 1), the indicator with the strongest predictive performance in each group in the first cycle data set was V_HBsAg 8W、V_HBsAb 8W 、V_ALT 8W , D_HBsAg 8W 、D_HBsAb8W、D_ALT 8W , P_HBsAg 8W 、P_HBsAb 8W and P_ALT 8W ; The indicator with the strongest predictive performance in each group in the first two cycles of the data set is V_HBsAg 8W 、V_HBsAb 8W 、V_ALT 8W , D_HBsAg 8W , D_HBsAb 8W 、D_ALT 8W , P_HBsAg 8W 、P_HBsAb 8W and P_ALT 8W ; The indicator with the strongest predictive performance in each group in the first three cycles of the data set is V_HBsAg 8W 、V_HBsAb 36W 、V_ALT 8W , D_HBsAg 8W , D_HBsAb 36W 、D_ALT 8W , P_HBsAg 36W 、P_HBsAb 36W and P_ALT 8W ; The indicator with the strongest predictive performance in each group in the first 4 cycles of the data set is V_HBsAg 48W 、V_HBsAb 48W 、V_ALT 8W , D_HBsAg 8W , D_HBsAb 48W 、D_ALT 8W , P_HBsAg 48W 、P_HBsAb 48W and P_ALT 8W ; The indicator with the strongest predictive performance in each group in the first 5 cycles of the data set is V_HBsAg 60W 、V_HBsAb 60W 、V_ALT 8W , D_HBsAg 8W , D_HBsAb 60W 、D_ALT 8W , P_HBsAg 60W 、P_HBsAb 60W and P_ALT 8W ; The indicator with the strongest predictive performance in each group in the first 6 cycles of the data set is V_HBsAg72W 、V_HBsAb 60W 、V_ALT 8W , D_HBsAg 8W , D_HBsAb 72W 、D_ALT 8W , P_HBsAg 72W 、P_HBsAb 72W and P_ALT 8W .

[0052] Table 1. Results of univariate logistic model regression parameters

[0053]

[0054]

[0055]

[0056] (3) Multivariate logistic regression screening of predictors using LASSO algorithm

[0057] The time-varying indicators obtained by grouped univariate logistic regression screening were combined with the non-time-varying indicators height, weight, age, sex, dose of pegbin, and V_HBsAg. 0W 、V_HBsAb 0W 、V_HBeAg 0W 、V_HBeAb 0W 、V_HBVDNA 0W and V_ALT 0W , perform multi-factor logistic regression of LASSO algorithm. Multi-factor logistic regression of LASSO algorithm shrinks the coefficient of candidate factors that contribute less to clinical endpoint prediction to 0 by adding penalty terms, thereby screening out key factors affecting clinical endpoint. The multi-factor logistic regression results of LASSO algorithm are shown in Table 2. Different key predictive factors were screened when predicting clinical endpoints in different treatment cycles, and the variance inflation factor (VIF) was less than 5, indicating that the method of screening predictive factors of the present invention avoids the multicollinearity problem of regression model.

[0058] Table 2 Regression parameter results of the prediction model

[0059]

[0060]

[0061] After multi-factor logistic regression using the LASSO algorithm, the key indicators obtained from the first cycle of the data set screening are Height, V_HBsAg 8W and P_HBsAg 8W , the scoring calculation formula is obtained by fitting the existing data:

[0062] L1=13.124-0.083×Height-0.002×V_HBsAg 8W -0.673×P_HBsAg 8W ;

[0063] The key indicators obtained by screening the datasets in the first two cycles are Height, V_HBsAg 8W and P_HBsAg 8W , and the scoring calculation formula is obtained:

[0064] L2=12.915-0.081×Height-0.002×V_HBsAg 8W -0.672×P_HBsAg 8W ;

[0065] The key indicator obtained by screening the first three cycles of data sets is P_HBsAg 36W and P_HBsAb 36W , and the scoring calculation formula is obtained:

[0066] L3=-2.873-0.539×P_HBsAg 36W +0.648×P_HBsAb 36W ;

[0067] The key indicator obtained by screening the first 4 cycles of data sets is D_HBsAb 48W , P_HBsAg 48W and P_HBsAb 48W , and the scoring calculation formula is obtained:

[0068] L4=-3.651-0.044×D_HBsAb 48W -0.739×P_HBsAg 48W -0.685×P_HBsAb 48W ;

[0069] The key indicator obtained by screening the first 5 cycles of data sets is P_HBsAg 60W and P_HBsAb 60W , and the scoring calculation formula is obtained:

[0070] L5=-3.862-0.616×P_HBsAg 60W +0.416×P_HBsAb60W ;

[0071] The key indicator obtained by screening the first 6 cycles of data sets is V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W , and the scoring calculation formula is obtained:

[0072] L6=-5.113+0.008×V_HBsAb 60W -0.811×P_HBsAg 72W +0.028×P_HBsAb 72W .

[0073] The above scoring formula is used as follows: select the corresponding formula based on the treatment cycle when the patient is tested. For example, if the patient is in the third cycle when tested, use the L3 scoring formula to calculate the patient's predicted score, and use the corresponding critical value of the formula / treatment cycle to determine whether the patient has HBsAg negative conversion at the end of treatment.

[0074] Example 2 Evaluation of the predictive performance of the model

[0075] Based on the modeling data, the ROC curve was used to evaluate the predictive efficacy of the six prediction models described in Example 1 on whether HBsAg conversion occurred after the end of treatment. The results showed that the areas under the ROC curves of prediction models L1, L2, L3, L4, L5 and L6 were 0.8286, 0.8264, 0.8924, 0.9224, 0.9222 and 0.9516 ( Figure 1 ), indicating that the prediction of the model can better distinguish patients who are ultimately HBsAg positive or negative, and the prediction efficiency of the model will gradually increase with the progress of the patient's treatment. The critical value of the conversion is determined by the point with the highest true positive rate and the lowest false positive rate on the ROC curve. The corresponding critical values ​​of the prediction models L1, L2, L3, L4, L5 and L6 are 0.3465, 0.3474, 0.2340, 0.2269, 0.4166 and 0.1936, respectively.

[0076] In addition, the present invention uses a calibration curve with 1000 resamplings and the Hosmer Lemeshow goodness of fit test to evaluate the consistency between the model prediction probability and the actual occurrence probability. The results show that the six prediction models can accurately predict the occurrence probability of HBsAg negative conversion in different patient subgroups from low to high probability, and the actual occurrence probability of the ordinate is in good agreement with the predicted probability of the abscissa ( Figure 2 ), all passed the Hosmer Lemeshow goodness-of-fit test (P=0.414, 0.470, 0.162, 0.627, 0.650, 0.907).

[0077] In order to fully examine the prediction ability of the model, we used the remaining 1 / 3 of the unmodeled data as external validation (validation data set). The validation data set was used to externally validate the above 6 models. The areas under the ROC curve of the prediction effectiveness of the prediction models from the first cycle to the first 6 cycles were 0.8174, 0.8174, 0.9162, 0.9480, 0.9385 and 0.9569 ( Figure 3 ), further indicating that the prediction of the aforementioned model can better distinguish patients who eventually undergo HBsAg negative conversion or not.

[0078] Similarly, the consistency between the model prediction probability and the actual occurrence probability was externally verified using the calibration curve of 1000 resamplings and the Hosmer Lemeshow goodness of fit test. The Hosmer Lemeshow goodness of fit test evaluation of the calibration curves drawn by the six prediction models based on the modeling data set (P = 0.482, 0.512, 0.291, 0.943, 0.597, 0.541) and the Hosmer Lemeshow goodness of fit test evaluation of the calibration curve after changing to the validation data set (P = 0.355, 0.572, 0.240, 0.478, 0.557, 0.524) all passed the statistical test, proving the consistency between the prediction probability obtained by the current method and the actual occurrence probability.

[0079] Example 3: Building a small program for dynamically predicting therapeutic effects and verifying the actual clinical test results

[0080] Based on the Shiny extension package of R language, six prediction models were integrated into a web applet that dynamically predicts whether patients will eventually become HBsAg negative during different treatment cycles of Peglib combined with NAs.

[0081] In the web applet, clinicians or patients can select the current cycle of Pegbin combined with NAs treatment and fill in the indicators required for prediction based on the treatment cycle data to calculate whether the patient can eventually achieve HBsAg negative conversion if he continues to receive the treatment. The web applet was used to test and clinically track 30 randomly selected patients enrolled in another clinical center. These 30 patients received the same treatment strategy as the 244 patients enrolled above.

[0082] Take one of the patients as an example to illustrate how to use the web app: This patient received Pegbin combined with NAs treatment for the 6th cycle and needed to enter the following indicators in the app:

[0083] (1) Baseline indicators (serum indicators obtained by blood test before patients received pegbin combined with NAs treatment): baseline HBsAg level 117.2 IU / mL, baseline HBsAb level 2.00 IU / mL, baseline HBeAg 0.083 COI, baseline HBeAb 0.008 COI;

[0084] (2) Indicators required in the L6 prediction model: HBsAg level 39.91 IU / mL at week 72, HBsAb level 2.00 IU / mL at week 60, and HBsAb level 2.00 IU / mL at week 72;

[0085] (3) Other patient data (optional): height 175 cm, pegol dose 180 μg.

[0086] By clicking the prediction button, the score of the prediction model is 0.0145. The conversion critical value corresponding to the prediction model L6 is 0.1936, which does not reach the threshold for HBsAg negative conversion. Therefore, the prediction model judges that the patient is still HBsAg positive after 12 cycles of peglibin combined with NAs treatment. The results of 12 cycles of clinical monitoring are consistent with the judgment results of the mini program.

[0087] The results of the webpage applet judgment and clinical actual monitoring of 30 patients at different treatment periods are shown in Table 3:

[0088] Table 3

[0089]

[0090]

[0091] According to the data in Table 3, the consistency rate between the prediction results of the mini program and the actual clinical follow-up results is 96.7%, indicating that the prediction results are highly accurate and can be used to help clinicians adjust the treatment plans of CHB patients and improve the treatment benefits of patients.

[0092] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A group of biomarkers for dynamically predicting the therapeutic efficacy of Peg-IFN-α combined with NAs, characterized in that: The biomarkers include Height, V_HBsAg 8W , P_HBsAg 8W , P_HBsAg 36W 、P_HBsAb 36W , D_HBsAb 48W , P_HBsAg 48W 、P_HBsAb 48W , P_HBsAg 60W 、P_HBsAb 60W 、V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W .

2. An evaluation model for dynamically predicting the therapeutic efficacy of Peg-IFN-α combined with NAs, characterized in that: The model uses the biomarker described in claim 1 as a key evaluation indicator.

3. The evaluation model according to claim 2, characterized in that: The model evaluates key indicators of the patient's samples according to the different treatment cycles of the patient, so as to predict the patient's treatment efficacy. The treatment is based on a 12-week treatment cycle. In one treatment cycle, Pegibine is treated for 8 weeks and then stopped for 4 weeks. NAs are continuously used for a total of 12 treatment cycles, including: The calculation formula for the prediction score of the first cycle is: L1=13.124-0.083×Height-0.002×V_HBsAg 8W -0.673×P_HBsAg 8W ; The prediction score calculation formula for the first two cycles is: L2=12.915-0.081×Height-0.002×V_HBsAg 8W -0.672×P_HBsAg 8W ; The prediction score calculation formula for the first three cycles is: L3=-2.873-0.539×P_HBsAg 36W +0.648×P_HBsAb 36W ; The prediction score calculation formula for the first 4 cycles is: L4=-3.651-0.044×D_HBsAb 48W -0.739×P_HBsAg 48W -0.685×P_HBsAb 48W ; The prediction score calculation formula for the first 5 cycles is: L5=-3.862-0.616×P_HBsAg 60W +0.416×P_HBsAb 60W ; The prediction score calculation formula for the first 6 cycles is: L6=-5.113+0.008×V_HBsAb 60W -0.811×P_HBsAg 72W +0.028×P_HBsAb 72W ; The evaluation model is used as follows: when the patient is in the nth treatment cycle (1<n≤6), the L calculated based on the corresponding key indicator data n The prediction score of L n The critical value is used to determine whether the patient has HBsAg negative conversion at the end of treatment.

4. The evaluation model according to claim 3, characterized in that: The corresponding critical values ​​of L1, L2, L3, L4, L5 and L6 are 0.3465, 0.3474, 0.2340, 0.2269, 0.4166 and 0.1936, respectively. If the value calculated by the model is higher than the corresponding critical value of the cycle, it is predicted that the patient will eventually be HBsAg positive (HBsAg has not turned negative), otherwise it is predicted that the patient will eventually be HBsAg negative (HBsAg has turned negative).

5. Use of the biomarker according to claim 1 in the preparation of a product for dynamically predicting the therapeutic efficacy of Peg-IFN-α combined with NAs, characterized in that: The products include test kits, chips, nucleic acid membrane strips, diagnostic equipment and diagnostic instruments.

6. A diagnostic system for dynamically predicting the therapeutic efficacy of Peg-IFN-α combined with NAs, characterized in that: The diagnostic system comprises: (1) Detection module: The detection module is used to detect the values ​​of multiple biomarkers, wherein the biomarkers are selected from Height, V_HBsAg 8W , P_HBsAg 8W , P_HBsAg 36W 、P_HBsAb 36W , D_HBsAb 48W , P_HBsAg 48W 、P_HBsAb 48W , P_HBsAg 60W 、P_HBsAb 60W 、V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W ; (2) Result judgment module: The result judgment module is used to determine the biomarkers Height, V_HBsAg detected by the detection module. 8W , P_HBsAg 8W , P_HBsAg 36W 、P_HBsAb 36W , D_HBsAb 48W , P_HBsAg 48W 、P_HBsAb 48W , P_HBsAg 60W 、P_HBsAb 60W 、V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W The value of the test subject is determined and the diagnostic model according to claim 3 outputs whether the test subject is ultimately HBsAg positive or negative.

7. The diagnostic system according to claim 6, characterized in that The working process of the user submission module includes the following: (1) The user fills in a form on the WeChat applet or web page, where the content is selected from Height, V_HBsAg 8W , P_HBsAg 8W , P_HBsAg 36W 、P_HBsAb 36W , D_HBsAb 48W , P_HBsAg 48W 、P_HBsAb 48W , P_HBsAg 60W 、P_HBsAb 60W 、V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W ; Fill in the indicator values ​​and patient baseline values ​​required for the score calculation formula for each cycle according to the patient's different treatment cycles: (2) The front end (mini program or web page) obtains the form data filled out by the user and sends the data to the back end program through a server request. The back end program obtains the user's diagnostic data by receiving HTTP requests and passes it into the diagnostic algorithm.

8. An electronic device, characterized in that: The electronic device comprises: (1) Input unit: The input unit is used to input multiple physiological index data of the subject, and the physiological index data is selected from Height, V_HBsAg 8W , P_HBsAg 8W , P_HBsAg 36W 、P_HBsAb 36W , D_HBsAb 48W , P_HBsAg 48W 、P_HBsAb 48W , P_HBsAg 60W 、P_HBsAb 60W 、V_HBsAb 60W , P_HBsAg 72W and P_HBsAb 72W ; (2) an analysis unit: the analysis unit is used to analyze and calculate the physiological indicator data input by the input unit according to the diagnostic model described in claim 3, and obtain the prediction score of the subject, and further predict whether the subject will have HBsAg negative conversion at the end of treatment according to the critical value described in claim 4; (3) Output unit: Outputs the results predicted by the analysis unit through screen display, tactile feedback, color feedback, etc.

9. The electronic device according to claim 8, characterized in that: The subject's physiological index data to be entered is selected based on the subject's treatment cycle.

10. A computer-readable storage medium, characterized in that: It stores a computer program, which is a computer program that realizes the function of the evaluation model as described in any one of claims 2 to 4 or the diagnostic system as described in claim 6 or 7.

Images