Application of reagent for detecting GDF15 protein in preparation of product for evaluating myocardial infarction risk of subject in future

Abstract

The invention provides application of a reagent for detecting GDF15 protein in preparation of a product for evaluating the myocardial infarction risk of a subject in the future, and belongs to the technical field of biomedical detection. By detecting the concentration level of the GDF15 protein in peripheral blood, the risk of myocardial infarction of a subject in the future can be accurately predicted in the early stage, so that early risk guidance is provided for crowds needing intervention, and the application is non-invasive, simple, convenient and low in cost. In order to achieve the maximum economic benefit, only the GDF15 level in the peripheral trace blood of the subject needs to be detected, the clinical index level needed in a traditional model is not needed, the prediction level of a traditional method is exceeded, and meanwhile the economic cost is greatly reduced.

Description

Technical Field

[0001] The present invention belongs to the field of biomedical detection technology, and in particular relates to the use of a reagent for detecting GDF15 protein in the preparation of a product for assessing the risk of a subject suffering from myocardial infarction in the future. Background Art

[0002] Cardiovascular disease, especially ischemic heart disease, is the leading cause of death worldwide and significantly increases the risk of disability. Despite improvements in reperfusion therapy and prognosis, myocardial infarction still leads to various complications, including cardiogenic shock and heart failure, which illustrates the importance of identifying risk factors for early prevention. Current risk assessment tools, such as the Framingham Risk Score and the Systematic Coronary Artery Risk Evaluation (SCORE2) system, have the potential to estimate the risk of developing coronary heart disease within 10 years. However, these tools rely solely on traditional risk factors and are limited in accurately predicting future myocardial infarction events due to the complex interactions of factors such as blood pressure, smoking, and cholesterol levels. In addition, these tools may not take into account the intricate and shared pathophysiological processes behind myocardial infarction.

[0003] Growth Differentiation Factor 15 (GDF15) is a protein produced by multiple organs such as the liver, intestines, muscles, and adipose tissue. It is a member of the TGF-beta superfamily and plays a role in a variety of physiological and pathological processes, including regulating appetite, weight control, inflammatory response, and cellular stress response. It has been reported in the past that it is closely related to obesity, diabetes, etc. GDF15 in plasma may be used as a molecular marker for disease. However, current research lacks discussion on the prediction time. There are still no reports on the effect of GDF15 alone in predicting a single long-term endpoint (such as myocardial infarction). Summary of the invention

[0004] In view of this, the object of the present invention is to provide a reagent for detecting GDF15 protein for use in the preparation of a product for assessing the risk of myocardial infarction in a subject in the future.

[0005] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions:

[0006] The present invention provides an application of a reagent for detecting GDF15 protein in the preparation of a product for assessing the risk of a subject suffering from myocardial infarction in the future.

[0007] Preferably, the reagent for detecting GDF15 protein includes a reagent for detecting the expression amount of GDF15 protein.

[0008] Preferably, the reagent comprises an antibody against the GDF15 protein.

[0009] Preferably, the reagent comprises an enzyme-linked immunosorbent assay kit.

[0010] Preferably, the future time is more than 10 years.

[0011] Preferably, the sample from the subject comprises peripheral blood.

[0012] Preferably, the product comprises a reagent or a kit.

[0013] Preferably, the evaluation standard is that when the NPX value of GDF15 is ≥ 0.542, the risk of the subject suffering from myocardial infarction in the future is high, and when the NPX value of GDF15 is < 0.542, the risk of the subject suffering from myocardial infarction in the future is low.

[0014] The present invention provides a system for assessing a subject's future risk of myocardial infarction, comprising a detection module and a judgment module, wherein the detection module is used to obtain the GDF15 protein concentration in plasma and convert the GDF15 protein concentration into a GDF15 protein NPX value, and the judgment module comprises a readable carrier storing a judgment standard.

[0015] Preferably, the judgment standard is that when the NPX value of GDF15 is ≥ 0.542, the risk of the subject suffering from myocardial infarction in the future is high, and when the NPX value of GDF15 is < 0.542, the risk of the subject suffering from myocardial infarction in the future is low.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] The present invention provides an application of a reagent for detecting GDF15 protein in the preparation of a product for assessing the risk of myocardial infarction in the future of a subject. The present invention can accurately predict the risk of myocardial infarction in the future of a subject at an early stage by detecting the concentration level of GDF15 protein in peripheral blood, so as to provide early risk guidance for people who need intervention. The application is non-invasive, simple and low-cost. In order to achieve the greatest economic benefit, it is only necessary to detect the level of GDF15 in the peripheral trace blood of the subject, without the clinical indicator level required in the traditional model, which not only exceeds the prediction level of the traditional method, but also greatly reduces the economic cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a protein spectrum with a high risk of myocardial infarction screened using the Cox risk regression model;

[0019] Figure 2 ROC curve of the lightGBM machine learning model built for the GDF15 protein;

[0020] Figure 3To obtain the survival results of the high-risk group and low-risk group samples using Kaplan-Meier survival analysis, the green line represents the low-risk group and the red line represents the high-risk group;

[0021] Figure 4 ROC curve of the lightGMB machine learning model constructed for GDF15 protein and the traditional clinical indicator risk model. DETAILED DESCRIPTION

[0022] The present invention analyzes the results of the highly sensitive Olink proteomics of the world's largest natural population cohort to date (a total of 2,923 proteins in more than 50,000 people were measured, with a follow-up period of about 14 years), and uses the highly efficient lightGBM machine learning model to mine plasma protein molecular markers, maximizing the rational use of existing data, thereby achieving a high prediction accuracy for the occurrence of myocardial infarction (such as ST-segment elevation myocardial infarction and non-ST-segment elevation myocardial infarction) in the future (more than 10 years).

[0023] In the early stage of screening protein markers, the present invention incorporates different personal background characteristics, clinical characteristics or proteomic data for machine learning testing to balance the prediction accuracy and the number of included indicators. The Cox risk regression model was used to screen the protein spectrum related to myocardial infarction from 2923 proteins, and then the machine learning algorithm lightGBM was used to rank the importance of proteins in the prediction of myocardial infarction. GDF15 was the highest ranked protein, and the lightGMB machine learning model constructed by GDF15 single protein had higher prediction performance for GDF15 protein than the model constructed by traditional clinical indicators.

[0024] Based on this, the present invention provides an application of a reagent for detecting GDF15 protein in the preparation of a product for assessing a subject's future risk of myocardial infarction.

[0025] In the present invention, the reagent for detecting GDF15 protein preferably includes a reagent for detecting the expression of GDF15 protein, and further preferably includes an antibody for GDF15 protein, and further preferably includes an enzyme-linked immunosorbent assay kit. The enzyme-linked immunosorbent assay kit is preferably an Olink kit, and the Olink kit is purchased from Thermo Fisher Scientific, product name Explore 3072 / 384. The present invention can be used to assess the risk of myocardial infarction in a subject in the future by detecting the concentration of GDF15 protein, wherein the future time is more than 10 years, such as 10 years, 11 years, 12 years, 13 years, 14 years or 15 years.

[0026] In the present invention, the sample of the subject includes peripheral blood, and detecting the GDF15 protein concentration in the plasma is a non-invasive, simple and low-cost method for evaluating the risk of myocardial infarction in the subject in the future. The product preferably includes a reagent or a kit. The evaluation standard is that when the NPX value of GDF15 is ≥0.542, the risk of myocardial infarction in the subject in the future is high, and when the NPX value of GDF15 is <0.542, the risk of myocardial infarction in the subject in the future is low.

[0027] The present invention provides a system for assessing a subject's future risk of myocardial infarction, comprising a detection module and a judgment module, wherein the detection module is used to obtain the GDF15 protein concentration in plasma and convert the GDF15 protein concentration into a GDF15 protein NPX value, and the judgment module comprises a readable carrier storing a judgment standard.

[0028] In the present invention, the judgment standard is that when the NPX value of GDF15 is ≥ 0.542, the risk of the subject suffering from myocardial infarction in the future is high, and when the NPX value of GDF15 is < 0.542, the risk of the subject suffering from myocardial infarction in the future is low.

[0029] In the present invention, unless otherwise specified, all raw material components are commercially available products well known to those skilled in the art.

[0030] The technical solutions provided by the present invention are described in detail below in conjunction with the embodiments, but they should not be construed as limiting the protection scope of the present invention.

[0031] In the following embodiments, the present invention has no particular limitation on the Olink Explore proximity extension analysis method, and any method known in the art may be used, such as the method described in document DOI10.1038 / s41586-023-06592-6.

[0032] The present invention uses an Olink kit to measure protein, and after obtaining the sample protein level, the sample protein level is converted into a normalized protein expression (NPX) value. The present invention has no particular limitation on the calculation method of the NPX value, and a method known in the art can be used, such as the method described in https: / / biobank.ndph.ox.ac.uk / showcase / ukb / docs / Olink_1536_B0_to_B7_Normalization.pdf (UKB-Olink Explore 1536-DataNormalizationStrategy).

[0033] Example 1

[0034] Clinical data, basic information, and omics data of 51,294 subjects from UKBiobank with Olink proteomics measurement data were analyzed.

[0035] (1) Peripheral blood samples collected using EDTA tubes were centrifuged at 2500 g for 10 minutes at 4°C to separate plasma, and the plasma was then aliquoted and stored at -80°C.

[0036] (2) The obtained plasma samples were subjected to protein quantification using the Olink Explore proximity extension assay, i.e., 2923 unique proteins were detected using the Olink kit to obtain the protein levels of 2923 samples, which were then converted into standardized protein expression (NPX) values.

[0037] (3) The NPX values ​​of 2923 proteins were used to construct a Cox risk regression model, where the outcome of the Cox risk regression model was a binary outcome: whether myocardial infarction occurred, and the variables were the NPX values ​​of 2923 proteins + a series of calibrated covariates (sex, age, education level, body mass index (BMI), high-density lipoprotein (HDL), low-density lipoprotein (LDL), smoking, total triglycerides (TG), total cholesterol (TC), systolic blood pressure (SBP), diastolic blood pressure (DBG), estimated glomerular filtration rate (eGFR), whether diabetes occurred, and C-reactive protein level). Cox risk regression model = coxph (surv_obj (annual survival rate of surviving subjects) ~ group (risk group) + age (age) + sex (sex) + edu (years of education), data = mydf (Olink proteomics dataset)). After the regression risk model was constructed, the coef (cox_model) function was used to obtain the risk coefficient of the risk model, which was the risk value. Cox risk regression was performed on each of the 2923 proteins to obtain the p-value of each single protein. The p-value was then corrected for multiple tests and proteins with p<0.05 were screened to obtain a protein spectrum with a high risk of myocardial infarction.

[0038] Figure 1 The results showed that 270 proteins were identified in the protein spectrum with a high risk of myocardial infarction and were associated with myocardial infarction outcomes, among which the risk value (HR) of GDF15 was 1.56 and the p value was 3.37×10 -27 .

[0039] (4) The protein spectra associated with a high risk of myocardial infarction were incorporated into the lightGBM machine learning model one by one, and the accuracy of the lightGBM machine learning model was evaluated by protein pair analysis.

[0040] lightGBM is an efficient machine learning algorithm based on the gradient boosting framework. It uses the gradient boosting framework to iteratively add decision trees to approach the optimal solution of the objective function. By gradually and iteratively adding protein information, it continuously evaluates whether the model has achieved the optimal prediction effect.

[0041] Specifically, lightgbm (version 4.3.0) is used for model prediction. The model calculates the gradient (first-order derivative) and Hessian matrix (second-order derivative) of the loss function to find the optimal tree structure. For a given loss function L, the gradient g and Hessian h are:

[0042] L(y, Y) = (yY) 2 .

[0043]

[0044] Among them, y is the true value, Y is the predicted value, F(x i ) represents the predicted value of the current model on sample i.

[0045] After the above optimization, the lightGBM machine learning model is configured according to the following key parameters: specify the evaluation target as binary classification (objective = "binary"), the evaluation standard as AUC (metric = "auc"), perform data imbalance processing (is_unbalance = TRUE), the learning rate is 0.01 (learning_rate = 0.01), the maximum depth of the tree is 15 (max_depth = 15), the number of leaves is 10 (num_leaves = 10), the subsampling ratio of the data is 0.7 (subsample = 0.7), the subsampling ratio of the feature is 0.7 (colsample_bytree = 0.7), and the number of training rounds is 500 (nrounds = 500).

[0046] Next, the "Leave-One-Out" method was used for cross-validation: 51,294 subjects were classified into 10 regional categories, and an iteration was set for each region, in which the data of the region was used as a test set, and the data of the remaining nine regions were combined as a training set. The data consists of two parts: outcome variables and predictor variables. The outcome variable is whether myocardial infarction occurs, and the predictor variable is proteomic data. This will be repeated 10 times, and a different region will be selected as the test set each time. After all 10 iterations are completed, the performance of the model on the test set of each region is summarized, and the average performance index is calculated to establish the optimal LightGBM machine learning model.

[0047] The lightGBM machine learning model was used to dynamically obtain the prediction and test data of 51,294 cases of protein group blood. The prediction data refers to the simulated prediction of whether each individual has developed a myocardial infarction using the constructed lightGBM model and protein indicators; the test data refers to whether each individual has actually developed a myocardial infarction. The pROC (version 1.18.5) encapsulated in the R language program was used to calculate the receiver operating characteristic (ROC) and the area under the curve (AUC). Finally, a ROC curve evaluation model containing the true positive rate (TPR) and the false positive rate (FPR) was created to evaluate the average performance index. The specific steps are (where y is the true value and Y is the predicted value): ROC = roc(y, Y); AUC = auc(roc_res); TPR = roc_res$sensitivities; FPR = 1-roc_res$specificities.

[0048] Youden index is a statistical method used to evaluate the performance of a classifier. It can find the optimal classification threshold by maximizing the sum of sensitivity and specificity. The coords function encapsulated in the R language program is used to obtain the optimal threshold point from the ROC curve analysis based on the Youden index.

[0049] Through the analysis of the lightGBM machine learning model, GDF15 is the protein molecule with the highest score for the lightGBM machine learning model.

[0050] Figure 2 The results showed that the lightGBM machine learning model predicted an AUC value of 0.69. In addition, through ROC analysis, the Youden optimal threshold point of GDF15 was 0.542. The present invention uses this as the prediction threshold point for the future risk of myocardial infarction, that is, when the NPX value of GDF15 is ≥ 0.542, the subject has a high risk of myocardial infarction in the future, and when the NPX value of GDF15 is < 0.542, the subject has a low risk of myocardial infarction in the future.

[0051] In this embodiment, 51294 subjects were divided into high-risk group and low-risk group according to their plasma GDF15 levels at the time of first enrollment. Thereafter, the occurrence of myocardial infarction in the respondents was followed up every 2.5 years for a total of 15 years. According to the occurrence of myocardial infarction at each follow-up, the survival curves of the two risk groups were drawn. The high-risk group refers to a high risk of myocardial infarction in the future, and the low-risk group refers to a low risk of myocardial infarction in the future: peripheral blood plasma was collected, and the protein level of GDF15 was detected according to the Olink kit, and the protein level of GDF15 was converted into NPX value, and then the high-risk group and low-risk group were determined. The determination method was: when the NPX value of GDF15 was ≥0.542, the risk of myocardial infarction in the future of the subject was high, and when the NPX value of GDF15 was <0.542, the risk of myocardial infarction in the future of the subject was low. Then the survival rates of the high-risk group and low-risk group samples were analyzed separately using Kaplan-Meier survival analysis.

[0052] Figure 3 The results showed that the average risk of myocardial infarction in the high-risk group was 2.16 times higher than that in the low-risk group. The probability of long-term myocardial infarction in the high-risk group was significantly higher than that in the low-risk group. The statistical test was significant, indicating that the use of GDF15 was a significant risk indicator for myocardial infarction in the subjects within the next 15 years.

[0053] Furthermore, in order to prove the high accuracy of using GDF15 to assess the risk of myocardial infarction in the next 10 years for subjects, the GDF15 risk model for each year before the prediction (i.e., the lightGMB machine learning model constructed above) was compared with the existing clinical indicator risk model. The clinical indicator risk model uses the recognized clinical predictive indicators of myocardial infarction: gender, age, smoking, systolic blood pressure and low-density lipoprotein as predictive variables. In addition, the construction of the clinical indicator risk model refers to step (4) of this embodiment. The difference between the two models for each year was subjected to a Delong test. If the p-value is less than 0.05, the difference between the models is considered to be significant.

[0054] Figure 4 The results showed that the single GDF15 protein was superior to the traditional indicators (gender, age, smoking, systolic blood pressure, non-high-density lipoprotein) in predicting the risk of future myocardial infarction. Ten years before the subjects had myocardial infarction, GDF15 was higher than traditional clinical indicators (AUC 0.69 vs. 0.536).

[0055] Example 2

[0056] The expression level of GDF15 protein is used to assess the risk of myocardial infarction in the future

[0057] 1.1 Study subjects and clinical data

[0058] This example included a total of 100 subjects with an average age of 60.2 years who were admitted to our hospital between 2005 and 2012, including 55 males and 45 females. All subjects had not experienced myocardial infarction. The occurrence of myocardial infarction within 10 years was continuously followed up.

[0059] 1.1.1 The diagnosis of myocardial infarction is made jointly by two cardiologists with more than 10 years of clinical experience. The main basis includes the following aspects:

[0060] (1) Biochemical indicators of myocardial injury:

[0061] A pattern of increases and decreases in serum cardiac markers such as troponin, especially troponin levels exceeding the upper limit of the 99th percentile reference value.

[0062] (2) Clinical manifestations:

[0063] Typical angina or precordial pain that lasts for more than 20 minutes and is not completely relieved by rest or nitroglycerin.

[0064] It may be accompanied by other symptoms such as sweating, nausea, vomiting, and difficulty breathing.

[0065] (3) Electrocardiogram (ECG) changes:

[0066] T segment elevation or depression, new left bundle branch block (LBBB), pathological Q wave, etc.

[0067] Dynamic T wave changes.

[0068] (4) Imaging evidence:

[0069] Echocardiography, cardiac magnetic resonance imaging (MRI), etc. showed new local ventricular wall motion abnormalities.

[0070] 1.1.2 Exclusion criteria include:

[0071] (1) Non-cardiac chest pain:

[0072] Chest pain caused by other reasons (such as gastroesophageal reflux disease, pleurisy, etc.).

[0073] (2) Other major diseases:

[0074] Recent history of severe heart disease (such as heart failure, heart surgery).

[0075] Active infection or other systemic disease (such as cancer, severe liver or kidney dysfunction).

[0076] (3) Medication or therapeutic intervention:

[0077] Patients who have recently received anticoagulant therapy or have a bleeding tendency.

[0078] Patients who have received thrombolysis or coronary intervention.

[0079] 1.2 Sample collection

[0080] Peripheral blood of the above 100 subjects was collected using EDTA tubes, and centrifuged at 2500 g for 10 minutes at 4°C to separate plasma.

[0081] 1.3 Detection of GDF15 concentration in plasma using Olink kit

[0082] By using Olink kit (purchased from Thermo Fisher Scientific, product name Explore3072 / 384) detected the GDF15 concentration in the subject's plasma, and calibrated the measured GDF15 concentration using the NPX method and converted it into the NPX value to obtain the NPX value of GDF15 to ensure the accuracy of the results.

[0083] 1.4 Criteria

[0084] When the NPX value of GDF15 is ≥0.542, the subject has a high risk of suffering from myocardial infarction in the future; when the NPX value of GDF15 is <0.542, the subject has a low risk of suffering from myocardial infarction in the future.

[0085] 1.5 Data analysis and statistics using NPX values ​​of GDF15

[0086] The subjects are divided into a high-risk group and a low-risk group according to their risk of future myocardial infarction. The high-risk group refers to subjects with a high risk of future myocardial infarction, and the low-risk group refers to subjects with a low risk of future myocardial infarction.

[0087] According to the above criteria, 35 of the 100 subjects were divided into the high-risk group and 65 into the low-risk group.

[0088] Follow-up and data collection: All subjects were followed up for 10 years, and the occurrence of myocardial infarction during this period was recorded.

[0089] Data results:

[0090] During the follow-up period, 4 patients in the high-risk group developed myocardial infarction, while no patients in the low-risk group developed myocardial infarction.

[0091] Statistical analysis: The chi-square test was used to calculate the P value and analyze the association between GDF15 levels and myocardial infarction risk. The p value was 0.013, indicating that GDF15 has a significant risk prediction effect. Since the number of patients in the low-risk group was 0, the OR value was infinite.

[0092] The results showed that the above criteria can be used to distinguish the risk of myocardial infarction in the next 10 years by detecting GDF15 concentration. The risk of myocardial infarction in the next 10 years in the high-risk group was significantly higher than that in the low-risk group. This finding shows the value of GDF15 as a potential biomarker for myocardial infarction risk assessment and supports its application in cardiovascular disease prevention strategies.

[0093] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. Use of a reagent for detecting GDF15 protein in the preparation of a product for assessing a subject's future risk of myocardial infarction.

2. The use according to claim 1, characterized in that: The reagent for detecting the GDF15 protein includes a reagent for detecting the expression amount of the GDF15 protein.

3. The use according to claim 2, characterized in that: The reagents include antibodies against GDF15 protein.

4. The use according to claim 3, characterized in that: The reagents include an enzyme-linked immunosorbent assay kit.

5. The use according to claim 1, characterized in that: The future time is more than 10 years.

6. The use according to claim 1, characterized in that: The sample from the subject includes peripheral blood.

7. The use according to claim 1, characterized in that: The product comprises a reagent or a kit.

8. The use according to claim 1, characterized in that: The evaluation standard is that when the NPX value of GDF15 is ≥ 0.542, the subject has a high risk of future myocardial infarction, and when the NPX value of GDF15 is < 0.542, the subject has a low risk of future myocardial infarction.

9. A system for assessing a subject's risk of future myocardial infarction, characterized in that: It comprises a detection module and a judgment module, wherein the detection module is used to obtain the GDF15 protein concentration in plasma and convert the GDF15 protein concentration into the GDF15 protein NPX value, and the judgment module comprises a readable carrier storing the judgment standard.

10. The system according to claim 9, characterized in that The judgment standard is that when the NPX value of GDF15 is ≥ 0.542, the subject has a high risk of future myocardial infarction, and when the NPX value of GDF15 is < 0.542, the subject has a low risk of future myocardial infarction.

Images