Protein markers for predicting the prognosis risk of idiopathic dilated cardiomyopathy patients and application thereof

Abstract

The application relates to a plasma protein marker for predicting the prognosis risk of patients with idiopathic dilated cardiomyopathy and application thereof. Through systematic analysis of plasma samples of patients with idiopathic dilated cardiomyopathy, five proteins, MYL3, ANP, ALDH6A1, PPIA and PKIA, are established as core markers closely related to the prognosis of patients. Based on the expression levels of the above protein markers, a combined evaluation model for predicting the risk of adverse cardiovascular events is constructed, which can effectively stratify the risk of composite endpoint events such as cardiovascular death, heart failure rehospitalization, left ventricular assist device implantation or heart transplantation. The method adopts plasma sample detection, has non-invasiveness, stability and good clinical accessibility, and can provide a reliable technical means for the prognosis risk evaluation and individualized management of patients with idiopathic dilated cardiomyopathy.

Description

Technical Field

[0001] This application belongs to the field of medical diagnostic reagents and biomarkers, specifically relating to a protein biomarker for predicting the prognostic risk of patients with idiopathic dilated cardiomyopathy and its application in prognostic assessment products. Background Technology

[0002] Idiopathic dilated cardiomyopathy (iDCM) is a common non-ischemic cardiomyopathy with significant heterogeneity in disease progression and clinical outcomes. Although some patients have relatively stable conditions with standard treatment, a considerable proportion progress to end-stage heart failure, experiencing adverse endpoint events such as recurrent hospitalizations for heart failure, cardiovascular death, left ventricular assist device implantation, or heart transplantation. Therefore, accurate prognostic risk assessment and early risk stratification for iDCM patients are significant challenges in current clinical management.

[0003] Cardiac remodeling is the core pathological basis for the occurrence and progression of iDCM, and its severity is closely related to patient prognosis. Multidimensional structural and functional changes, including left ventricular dilation and decreased systolic function, right ventricular dysfunction, atrial dysfunction, and myocardial fibrosis, have all been proven to be significantly associated with adverse cardiovascular outcomes. However, current prognostic assessment methods largely rely on cardiac imaging examinations. While cardiac magnetic resonance imaging (MRI) has advantages in structural, functional, and histological assessment, its widespread and dynamic clinical application is limited by cost, equipment accessibility, and follow-up convenience. Meanwhile, conventional echocardiography still has limited ability to characterize complex cardiac remodeling changes and their potential prognostic significance, making it difficult to identify some high-risk patients in a timely manner.

[0004] In recent years, plasma proteomics has provided a new technical approach for the prognostic assessment of cardiovascular diseases. Blood samples are readily available and can reflect various pathophysiological processes associated with the progression and adverse outcomes of intravascular coagulation (iDCM), such as inflammatory responses, metabolic disorders, extracellular matrix remodeling, and myocardial injury, giving them potential advantages as a prognostic risk assessment tool. However, systematic studies on the use of plasma protein biomarkers in predicting the prognostic risk of iDCM are still relatively limited. Stable predictive models and application protocols based on combinations of multiple protein biomarkers and validated by clinical outcomes are lacking, making it difficult to meet the practical needs of precise clinical risk stratification. Summary of the Invention

[0005] Based on this, one embodiment of this application provides a blood protein biomarker for predicting the prognostic risk of patients with idiopathic dilated cardiomyopathy, and proposes its application in the preparation of products for prognostic risk assessment of patients with idiopathic dilated cardiomyopathy.

[0006] The technical solutions include the following:

[0007] In a first aspect, this application provides a protein biomarker for predicting the prognostic risk of patients with idiopathic dilated cardiomyopathy, said protein biomarker comprising one or more of the following: Aldehyde Dehydrogenase 6 Family Member A1 (ALDH6A1), Myosin Light Chain 3 (MYL3), Peptidylprolyl Isomerase A (PPIA), Protein Kinase Inhibitor Alpha (PKIA), and A-type Natriuretic Peptide (ANP).

[0008] In one embodiment, compared with patients with idiopathic dilated cardiomyopathy who had a lower prognostic risk, patients with a higher prognostic risk had elevated levels of MYL3 and ANP in their plasma samples, while the levels of ALDH6A1, PPIA, and PKIA were decreased.

[0009] In one embodiment, the sample comprises plasma.

[0010] Secondly, this application provides the application of the detection reagent for the expression level of the above-mentioned protein biomarkers in the preparation of an assessment product for predicting the prognostic risk of patients with idiopathic dilated cardiomyopathy.

[0011] In one embodiment, the product is used to predict the risk of a composite endpoint event, namely cardiovascular death, readmission from heart failure, implantation of a left ventricular assist device, or heart transplantation.

[0012] In one embodiment, the detection reagent is selected from one or more of the following methods: liquid chromatography, mass spectrometry, and immunoassay.

[0013] In one embodiment, the immunoassay includes one or more of Western blot analysis, radioimmunoassay, immunofluorescence assay, immunoprecipitation, immunodiffusion, electrochemiluminescence immunoassay, ELISA assay, and immunopolymerase chain reaction.

[0014] In one embodiment, the detection reagent is a reagent suitable for liquid chromatography-tandem mass spectrometry (LC-MS / MS) or double antibody sandwich ELISA.

[0015] In one embodiment, mass spectrometry based on protein characteristic tag peptides is employed.

[0016] In one embodiment, the detection reagent comprises an antibody or a functional fragment thereof.

[0017] In one embodiment, the product includes a reagent kit, a chip, a test strip, a system, or a device.

[0018] In one embodiment, the kit is a mass spectrometry identification tag peptide kit or an enzyme-linked immunosorbent assay (ELISA) kit.

[0019] In one embodiment, mass spectrometry identification of the tagged peptide refers to its use in a data-independent mass spectrometry acquisition mode.

[0020] Thirdly, this application provides a method for constructing a prognostic risk scoring model for idiopathic dilated cardiomyopathy, including construction using the expression levels of protein markers MYL3, ANP, ALDH6A1, PPIA, and PKIA.

[0021] In one embodiment, the prognostic risk scoring model is constructed using Cox regression based on expression data of the protein biomarkers in patients with idiopathic dilated cardiomyopathy.

[0022] Fourthly, this application provides a prognostic risk scoring model for patients with idiopathic dilated cardiomyopathy, the prognostic risk scoring model including MYL3, ANP, ALDH6A1, PPIA, and PKIA protein markers; optionally, the prognostic risk scoring model calculates the risk score using the following formula:

[0023] Risk Score = 0.4645 × ANP + 0.1174 × MYL3 - 0.3182 × PPIA - 0.1580× PKIA - 0.1516 × ALDH6A1

[0024] Wherein, Risk Score represents the risk score, ANP represents the expression level of ANP protein, MYL3 represents the expression level of MYL3 protein, PPIA represents the expression level of PPIA protein, PKIA represents the expression level of PKIA protein, and ALDH6A1 represents the expression level of ALDH6A1 protein.

[0025] The prognostic risk scoring model constructed based on the protein biomarkers calculates the risk score of patients to experience multiple adverse events, which can effectively distinguish between high-risk and low-risk patients. Patients in the high-risk group have a significantly increased risk of cardiovascular death, heart failure rehospitalization, left ventricular assist device implantation, or heart transplantation.

[0026] In one embodiment, a patient is identified as high-risk when the Risk Score is ≥ -0.469, and as low-risk when the Risk Score is < -0.469.

[0027] Fifthly, this application provides a prediction system or device for the prognostic risk of idiopathic dilated cardiomyopathy, the prediction system or device comprising the following modules (1) to (3):

[0028] (1) Data acquisition module, used to acquire the expression level of protein biomarkers in the sample of the subject to be tested, wherein the protein biomarkers include one or more of MYL3, ANP, ALDH6A1, PPIA and PKIA.

[0029] (2) Analysis module, used to analyze the expression levels of protein biomarkers obtained by the data acquisition module in order to assess the prognostic risk level of the subjects.

[0030] In one embodiment, the analysis module is used to input the expression level of the protein biomarker as input data into a pre-constructed prognostic risk scoring model to obtain the patient's prognostic risk score and risk stratification results.

[0031] In one embodiment, the prognostic risk scoring model is constructed based on the expression levels of the protein biomarkers in patients with idiopathic dilated cardiomyopathy, and is modeled using the Cox regression algorithm or other machine learning algorithms.

[0032] It should be noted that, in addition to the Cox regression algorithm, other machine learning algorithms can also be used to model the expression levels of the protein biomarkers, including but not limited to linear regression, support vector machine, nearest neighbor or k-nearest neighbor, decision tree, random forest, Naive Bayes, and gradient enhancement algorithms. Any system that constructs a prognostic risk prediction model based on the expression levels of the protein biomarkers described in this application and is used for prognostic risk assessment of patients with idiopathic dilated cardiomyopathy falls within the scope of protection of this application.

[0033] Optionally, the prognostic risk scoring model is the prognostic risk scoring model described in the fourth aspect.

[0034] (3) Result output module, used to output the analysis results of the analysis module to obtain the prognostic risk score and risk stratification results of the subjects.

[0035] Sixthly, this application provides a computer-readable storage medium including a computer program, which is executed by a processor and, when executed by the processor, implements the functions of different modules in the system.

[0036] In a seventh aspect, this application provides an electronic device including a memory and a processor, the memory storing a computer program and the processor executing the computer program to implement the functions of the prediction system.

[0037] It should be understood that the terms "system," "apparatus," and / or "device" as used herein are used to distinguish different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions. Those skilled in the art will recognize that this application can be implemented as a method or a computer program product. Therefore, the disclosure of this application can be specifically implemented in the following forms: entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software. Furthermore, in some specific embodiments, this application can also be implemented as a computer program product contained in one or more computer-readable media, which contains computer-readable program code.

[0038] Based on quantitative proteomics methods, this application identified five protein biomarkers—MYL3, ANP, ALDH6A1, PPIA, and PKIA—significantly associated with adverse clinical outcomes in the plasma of patients with idiopathic dilated cardiomyopathy, and constructed a stable multi-protein combined prognostic risk scoring model.

[0039] Compared with traditional technologies, this application has the following advantages:

[0040] This application employs a mass spectrometry method based on characteristic protein-tagged peptides to detect the expression levels of MYL3, ANP, ALDH6A1, PPIA, and PKIA in plasma. Through combined analysis of these five protein biomarkers, the prognostic risk of patients with idiopathic dilated cardiomyopathy can be effectively assessed, and high-risk and low-risk patients can be distinguished, thus providing a novel non-invasive detection method for prognostic risk assessment and risk stratification of this disease.

[0041] This application provides a non-invasive, stable plasma testing protocol suitable for clinical translation, which helps to achieve early prognostic risk identification and individualized follow-up management for patients with idiopathic dilated cardiomyopathy, and provides a new technical approach for clinical decision-making. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application and to more completely understand this application and its beneficial effects, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a correlation heatmap; the figure shows the Spearman correlation coefficient matrix between 27 cardiovascular magnetic resonance (CMR) parameters used for principal component analysis, with color intensity indicating the magnitude of the correlation coefficient.

[0044] Figure 2 The graph is a scree plot for principal component analysis; the bar chart represents the proportion of variance explained by each principal component, and the line chart represents the proportion of variance explained cumulatively.

[0045] Figure 3 This is a heatmap of the loading values ​​of the CMR parameters after principal component analysis. The figure shows the distribution of the loading values ​​of the first four principal components (PC1–PC4) obtained after principal component analysis of 27 CMR parameters and orthogonal rotation by Varimax. The colors indicate the magnitude and direction of the loading values ​​of each parameter on the corresponding principal component.

[0046] Figure 4 Figure A shows the clustering results of patients based on risk scores; Figure B shows that the within-group sum of squares elbow method indicates that K=2 is the optimal number of clusters; Figure C shows that the silhouette coefficient analysis shows that the two types of patients have high consistency; Figure C shows that the kernel density distribution plot shows that the risk score distributions of the two types of patients are clearly distinguishable and have little overlap.

[0047] Figure 5 Figure A shows the distribution of the two types of patients on each principal component score, defined as the advanced remodeling subtype and the compensated function subtype, respectively. Figure B shows the Kaplan-Meier event-free survival curves for the two types of patients, with the horizontal axis representing follow-up time and the vertical axis representing survival probability.

[0048] Figure 6Figure 1 is a schematic diagram of the protein screening process. Figure A is a LASSO regression coefficient path diagram, showing the coefficient shrinkage process of 1,309 differentially expressed proteins under different log-transformation penalty parameters λ. The dashed line represents the optimal penalty parameter (λ-min) determined through cross-validation. Figure B is a 10-fold cross-validation curve, showing the changes in the mean squared error (MSE) of the model under different log(λ) values, where λ-min corresponds to the minimum cross-validation error. Figure C shows 20 candidate proteins screened after LASSO regression and correction for age, sex, and NT-proBNP. Their regression coefficients are displayed in order of absolute value. In the bar chart, blue and red represent proteins that are positively and negatively correlated with the outcome, respectively.

[0049] Figure 7 This is a graph showing the expression distribution of candidate proteins in different subtypes. The graph uses a violin plot to illustrate the expression levels of ALDH6A1, MYL3, PPIA, PKIA, and ANP in the two patient groups. Cluster 1 represents the more severe cardiac remodeling type, and cluster 2 represents the functional compensatory type.

[0050] Figure 8 Kaplan–Meier curves for risk scores constructed based on the expression levels of five proteins; Subjects were divided into high and low risk groups based on the median risk score. The horizontal axis represents the follow-up time, and the vertical axis represents the event-free survival probability. The hazard ratio (HR) and 95% confidence interval are marked on the graph.

[0051] Figure 9 Kaplan–Meier survival curves for the external validation cohort; the vertical axis represents the survival probability, the horizontal axis represents the follow-up time (months), and the number of people at risk at the corresponding time point is marked below the curve. The survival difference between the two cardiac remodeling phenotypes was assessed by the log-rank test.

[0052] Figure 10 The expression distribution of five plasma proteins in two patient groups is shown in a violin diagram. The expression levels of ALDH6A1, MYL3, PPIA, PKIA and ANP are shown in the form of a violin diagram. Cluster 1 represents the more severe cardiac remodeling and cluster 2 represents the functional compensatory type.

[0053] Figure 11Kaplan–Meier curves of the Cox regression risk score constructed based on the expression levels of five plasma proteins in an external validation cohort; subjects were divided into high and low risk groups according to the median risk score, with the horizontal axis representing follow-up time and the vertical axis representing event-free survival probability, and the log-rank test results are labeled. Detailed Implementation

[0054] To make the above-mentioned objectives, features, and advantages of this application more apparent and understandable, a detailed description of specific embodiments of this application is provided below. Many specific details are set forth in the following description to provide a thorough understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of this application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0055] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0056] In this document, the term "and / or" includes any and all combinations of one or more of the related listed items.

[0057] In this paper, the term "Harrell's C index" refers to a statistical metric used to evaluate the discriminative ability of a predictive model. It is mainly used in time-outcome data analysis and measures the consistency between the model's predicted risk ranking and the actual order of events. Specifically, Harrell's C index reflects the probability that the model can correctly distinguish the risk of an individual whose event occurred earlier than that of an individual whose event occurred later in any pair of comparable individuals. The higher the index value, the stronger the model's ability to distinguish individual outcome risks.

[0058] In this study, severe remodeling or a more severe state of cardiac remodeling refers to a subtype of idiopathic dilated cardiomyopathy patients identified with higher risk scores after dimensionality reduction analysis to construct a comprehensive risk score characterizing the state of cardiac remodeling based on cardiac structural and functional parameters obtained from cardiac magnetic resonance imaging (CMR) and subsequent classification using unsupervised clustering. This subtype exhibits characteristic patterns in the principal component score spectrum associated with aggravated cardiac structural remodeling and functional impairment, and demonstrates poorer clinical outcomes during follow-up. These patients present with significantly impaired overall systolic and dystrophic function, right atrial mechanical dysfunction, and a more severe degree of myocardial remodeling, resulting in a higher risk of adverse clinical outcomes.

[0059] In this study, "relatively compensated" or "mild cardiac remodeling" refers to a subtype of idiopathic dilated cardiomyopathy patients identified by dimensionality reduction analysis of cardiac structural and functional parameters obtained from cardiac magnetic resonance imaging (CMR) to construct a risk score comprehensively representing the state of cardiac remodeling, followed by unsupervised clustering. This subtype exhibits a characteristic pattern in the principal component score spectrum associated with relatively preserved cardiac function and mild remodeling, and demonstrates relatively stable clinical outcomes during follow-up. These patients show relatively preserved overall cardiac function and atrial biomechanics, mild myocardial remodeling, and a relatively good clinical prognosis.

[0060] In this article, the term "prognosis" refers to the prediction of disease progression and outcome based on clinical data, including two types: natural prognosis (the outcome without treatment) and interventional prognosis (the outcome after medical intervention).

[0061] In this article, the term "NT-proBNP" refers to N-terminal pro–B-type natriuretic peptide, a biomarker secreted by ventricular myocytes when cardiac pressure load increases. It reflects cardiac function and is widely used to assess the severity and prognostic risk of heart failure.

[0062] This application relates to five biomarkers derived from plasma samples: myosin light chain 3 (MYL3, UniProt ID: P08590), anaerobic peptide A (ANP, UniProt ID: P01160), aldehyde dehydrogenase 6 family member A1 (ALDH6A1, UniProt ID: Q02252), peptidylprolyl isomerase A (PPIA, UniProt ID: P62937), and protein kinase inhibitor alpha (PKIA, UniProt ID: P61925). Among them, MYL3 is an important structural protein of the myocardial contractile apparatus, and changes in its expression level can reflect the state of myocardial structure and contractile function; ANP is a cardiac-derived endocrine peptide hormone that is closely related to ventricular load and cardiac remodeling; ALDH6A1 participates in mitochondrial-related metabolic processes and plays a role in energy metabolism homeostasis and oxidative stress regulation; PPIA, as a molecular chaperone protein, participates in protein folding and cellular stress response; and PKIA participates in intracellular signal regulation by inhibiting the protein kinase A signaling pathway.

[0063] Currently, there are no publicly available documents or patent reports that utilize the expression levels of MYL3, ANP, ALDH6A1, PPIA, and PKIA in plasma, or models constructed using these in combination, for prognostic risk assessment in patients with idiopathic dilated cardiomyopathy.

[0064] This application utilizes data-independent acquisition (DIA) mass spectrometry analysis combined with multivariate statistical modeling to screen and establish five proteins—MYL3, ANP, ALDH6A1, PPIA, and PKIA—as core prognostic biomarkers from the plasma of patients with idiopathic dilated cardiomyopathy. Patients were followed up for a median of 17.1 months (interquartile range 8.5–26.3 months), and composite endpoint events including cardiovascular death, heart failure readmission, left ventricular assist device implantation, or heart transplantation were systematically collected. Based on the expression levels of these five protein biomarkers, a protein risk scoring model was constructed using a multivariate Cox regression model (adjusted for age, sex, and NT-proBNP), and a prognostic risk score was calculated for each patient. Patients were divided into high-risk and low-risk groups based on the median risk score (cut-off = -0.469). Kaplan–Meier survival analysis and Cox proportional hazards model analysis showed that the risk of the composite endpoint event was significantly higher in the high-risk group than in the low-risk group (HR = 3.95, 95% CI: 1.67–9.34, P = 0.001). Further statistical analysis showed that the protein risk scoring model had good discriminative ability in prognostic prediction, with a Harrell's C index of 0.79 (95% CI: 0.74–0.86).

[0065] The embodiments of this application will be described in detail below with reference to examples. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of this application. For experimental methods in the following embodiments where specific conditions are not specified, please refer to the guidelines given in this application, or follow experimental manuals or conventional conditions in the art, or follow the conditions recommended by the manufacturer, or refer to experimental methods known in the art.

[0066] In the specific embodiments described below, the measurement parameters involving raw material components may have slight deviations within the weighing accuracy range unless otherwise specified. Temperature and time parameters are subject to acceptable deviations due to instrument testing accuracy or operational precision.

[0067] Example 1

[0068] 1. Materials and Reagents

[0069] The main instruments used in this embodiment included an Orbitrap Astral high-resolution mass spectrometer (Thermo Fisher Scientific) and a Vanquish Neo ultra-high performance liquid chromatography system (Thermo Fisher Scientific). The magnetic bead-type plasma protein processing kit used in sample pretreatment was purchased from Shanghai Bopu Biotechnology Co., Ltd. Trypsin and Lys-C / trypsin complex enzyme were purchased from Promega. Ammonium bicarbonate, tris(2-carboxyethyl)phosphine (TCEP), chloroacetamide (CAA), and other reagents were purchased from Sigma-Aldrich. Acetonitrile, formic acid, and HPLC-grade water were purchased from JT Baker. All chemical reagents were of analytical grade or mass spectrometry grade.

[0070] 2. Sample Source

[0071] This study included 175 plasma samples, comprising 125 patients with idiopathic dilated cardiomyopathy (iDCM) and 50 healthy controls. All plasma samples from iDCM patients were collected at the time of initial diagnosis, while samples from healthy controls were collected from individuals undergoing routine physical examinations during the same period. After collection, whole blood samples were centrifuged to separate plasma, aliquoted, and immediately stored at -80°C to avoid repeated freeze-thaw cycles for subsequent proteomics analysis.

[0072] 3. Plasma sample processing and proteolytic digestion

[0073] Take 50 μL of plasma sample from each case and use the magnetic bead plasma protein processing kit according to the manufacturer's instructions. First, add 20 μL of magnetic beads to the washing buffer and wash three times, mixing at 1300 rpm for 30 seconds each time. After magnetic adsorption for 1 minute, discard the supernatant. Then, add 50 μL of plasma sample and 50 μL of incubation buffer to the magnetic beads, mix at 1300 rpm at room temperature for 10 minutes to allow plasma proteins to fully bind to the surface of the magnetic beads. After incubation, discard the supernatant by magnetic adsorption and wash four times consecutively with washing buffer to remove non-specific components.

[0074] In the elution step, 40 μL of elution buffer was added to the magnetic beads and mixed at room temperature for 20 minutes to obtain the eluted protein. Then, 1 μL of 400 mM TCEP and 2 μL of 800 mM CAA were added to the eluent, and the reaction was carried out at 95 °C for 3 minutes to complete the protein reduction and alkylation. After cooling, 160 μL of 50 mM ammonium bicarbonate buffer was added, along with 0.5 μg of Lys-C / trypsin complex enzyme, and the reaction was incubated overnight at 37 °C. The reaction was terminated by adding a stop solution. The digested peptide solution was transferred to a C18 StageTip for desalting, and washed sequentially with methanol and 0.1% trifluoroacetic acid solution, followed by elution with 50% acetonitrile solution containing 0.1% trifluoroacetic acid. The eluent was dried by vacuum centrifugation and then reconstituted in a buffer solution containing 2% acetonitrile and 0.1% formic acid for subsequent liquid chromatography-mass spectrometry analysis.

[0075] 4. LC-MS / MS Analysis of DIA Data Acquisition

[0076] Peptide samples were separated using a Vanquish Neo ultra-high performance liquid chromatography system and detected by an OrbitrapAstral mass spectrometer. Samples were loaded onto a 50 cm μPAC™ Neo column, with 0.1% formic acid aqueous solution as mobile phase A and 0.1% formic acid in 80% acetonitrile as mobile phase B, and separation was achieved under gradient elution conditions for 8 minutes. Mass spectrometry data were acquired in data-independent acquisition (DIA) mode, with a primary mass spectrometry scan range of m / z 380–980 and a resolution of 240,000; and a secondary mass spectrometry scan range of m / z 150–2000, an isolation window width of 2 m / z, and a cycle time of 0.6 seconds.

[0077] 5. Sequence database retrieval and quantitative analysis

[0078] The obtained raw DIA data were analyzed using DIA-NN (version 1.8.18) software, and the UniProtKB / Swiss-Prot human protein database (October 2023 version) was searched. Search criteria were set to trypsin digestion, allowing a maximum of one missed cleavage site, cysteine ​​alkylation as a fixed modification, and N-terminal acetylation and methionine oxidation as variable modifications. False positive rate (FDR) was controlled to be less than 1% for both peptide and protein levels.

[0079] 6. Differential protein screening and functional annotation

[0080] Protein quantification results were imported into R statistical software for differential expression analysis. Differentially expressed proteins between patients with idiopathic dilated cardiomyopathy and healthy controls were screened by setting a Fold change >1.5 or <0.68, and an FDR-corrected P-value <0.05. Based on these screening criteria, a total of 1,309 proteins with significant differential expression between the two groups were identified. Protein functional annotation and pathway enrichment analyses were performed using the UniProt, Gene Ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases.

[0081] 7. Construction of cardiac remodeling phenotypes based on cardiac magnetic resonance parameters

[0082] To construct an objective, reproducible, and clinically valuable classification standard for cardiac remodeling in patients with idiopathic dilated cardiomyopathy, this embodiment systematically models and defines the cardiac remodeling phenotype based on quantitative parameters from cardiac magnetic resonance imaging (CMR). Unlike previous methods that relied on single clinical indicators such as left ventricular ejection fraction or functional classification for classification, this embodiment directly constructs a remodeling phenotype system reflecting cardiac structure, function, and histological characteristics based on multidimensional CMR imaging parameters.

[0083] The cardiac remodeling classification criteria in this application consist of four key steps:

[0084] (1) Data input and preprocessing;

[0085] (2) Principal component analysis (PCA) and calculation of reconstructed axis scores;

[0086] (3) Construction of prognostic risk score based on reconstructed axis;

[0087] (4) K-means clustering classification based on risk scores.

[0088] All analyses were performed using the R programming language.

[0089] 7.1 Fixed set of input parameters and preprocessing

[0090] To comprehensively reflect the structural, functional, and myocardial histological characteristics of the heart, this application selects 27 quantitative CMR parameters as modeling inputs, specifically including:

[0091] (1) Left ventricular structural and functional parameters: left ventricular end-diastolic volume index (LVEDVI), left ventricular end-systolic volume index (LVESVI), left ventricular ejection fraction (LVEF), left ventricular mass index (LVMI), and left ventricular global radial strain (LVGRS), global circumferential strain (LVGCS) and global longitudinal strain (LVGLS).

[0092] (2) Structural and functional parameters of the right ventricle: right ventricular end-diastolic volume index (RVEDVI), right ventricular end-systolic volume index (RVESVI), right ventricular ejection fraction (RVEF), right ventricular mass index (RVMI), and right ventricular global radial strain (RVGRS), global circumferential strain (RVGCS), and global longitudinal strain (RVGLS);

[0093] (3) Left atrial phase function and volume parameters: left atrial maximum volume index (LAVImax), left atrial total emptying fraction (LAEFtotal), and left atrial reserve strain (εs-LA), conduction strain (εe-LA) and boost strain (εa-LA);

[0094] (4) Right atrial phase function and volume parameters: right atrial maximum volume index (RAVImax), right atrial total emptying fraction (RAEFtotal), right atrial reserve strain (εs-RA), conduction strain (εe-RA), and boost strain (εa-RA);

[0095] (5) Myocardial histology parameters: delayed enhancement mass (LGEmass), extracellular volume fraction (ECV), and T1 relaxation time before contrast agent enhancement (preT1).

[0096] Data preprocessing methods:

[0097] Import the input data into the R language dataframe Patient_CMR_Data as a CSV or Excel file.

[0098] After performing a log2 transformation on the skewed distributed variables (LGEmass, ECV, preT1), all variables are standardized.

[0099]

[0100] in: The original variable value, This represents the mean of the variable in the discovery queue. The standard deviation is denoted as .

[0101] After standardization, the mean of the variables is 0 and the variance is 1 to ensure the repeatability and dimensional consistency of subsequent analyses.

[0102] 7.2 PCA Dimensionality Reduction and Reconstruction Axis Construction

[0103] (1) Correlation analysis:

[0104] Use `R cor()` to calculate the pairwise Pearson correlation coefficients of the 27 parameters and plot a heatmap to show the relationships between the variables. Figure 1 ).

[0105] (2) Principal component analysis

[0106] Principal component analysis was performed on the standardized variables using the R function prcomp().

[0107] The number of principal components retained was determined based on the following criteria: a screen plot was plotted to show the proportion of variance explained by each principal component and the cumulative explained variance; the number of principal components to retain was determined when the cumulative explained variance was ≥70% and the curve showed an inflection point. Ultimately, 4 principal components were retained. Figure 2 ).

[0108] (3) Rotation optimization

[0109] To improve interpretability, the principal() function in the R psych package is used with rotate="varimax" to orthogonally rotate the loading matrix, resulting in a rotated loading matrix.

[0110] (4) Reconstructing axis definition:

[0111] In the discovery queue, the loading coefficients of each variable on the four principal components are extracted based on the above rotational loading results, forming a 27 × 4 loading coefficient matrix. This matrix reflects the contribution weight of each CMR parameter on each principal component and serves as the fixed weight basis for subsequent reconstruction axis calculations.

[0112] Based on the rotating load structure, the four principal components are defined as four cardiac remodeling axes ( Figure 3 )

[0113] Reconstruction axis 1 (PC1): left and right ventricular systolic function and strain parameters, reflecting the overall contraction-deformation state.

[0114] Remodeling axis 2 (PC2): Biventricular volume and myocardial mass parameters, reflecting the degree of ventricular structural remodeling.

[0115] Reconstruction axis 3 (PC3): Right atrial volume and strain parameters, reflecting the mechanical functional state of the right atrium.

[0116] Reconstructed axis 4 (PC4): ECV and preT1 parameters, reflecting the level of diffuse myocardial fibrosis.

[0117] (5) Calculation of reconstructed axis score:

[0118] The score for each patient on the k-th reconstruction axis is calculated using the following formula:

[0119]

[0120] in Let i be the rotational load on the principal component k. These are the standardized input variables.

[0121] The PC value for each patient is automatically generated by predict(prcomp_model).

[0122] 7.3 Construction of Prognostic Risk Score Based on Reconstruction Axis

[0123] (1) Multivariate Cox model

[0124] A multivariate Cox proportional hazards model was constructed using the `coxph()` function from the `survival` package in the R language, with PC1–PC4 included as continuous predictor variables for model fitting. β1, β2, β3, and β4 are the regression coefficients (log hazard ratios) of PC1–PC4 in the Cox model, obtained through partial likelihood estimation. These regression coefficients were used to construct individualized prognostic risk scores.

[0125] (2) Calculation of individualized risk score

[0126] Calculate the prognostic risk score (RiskScore) for each patient based on the Cox regression coefficients β1–β4:

[0127]

[0128] The RiskScore is used to quantify the overall reconstruction risk for each patient, providing input features for subsequent clustering.

[0129] 7.4 K-means Clustering Based on Risk Scores

[0130] (1) Clustering input

[0131] Unsupervised clustering is performed using RiskScore as the input feature.

[0132] Since RiskScore is a continuous variable, the K-means algorithm is used to automatically determine the fractal boundary based on minimizing the intra-cluster squared error, rather than manually setting a threshold.

[0133] (2) Determining the optimal number of clusters

[0134] The optimal number of categories is determined using the following method:

[0135] Elbow method: Comparing the trend of changes in the sum of squares within clusters

[0136] Silhouette coefficient: assesses intra-class consistency

[0137] The optimal number of clusters was finally determined to be K=2 ( Figure 4 AB).

[0138] (3) Cluster stability verification

[0139] Kernel density estimation (KDE) was used to observe intra-cluster concentration and inter-cluster separation. Figure 4 C), verify the stability of the fractal.

[0140] (4) Classification results

[0141] Two stable cardiac remodeling phenotypes were ultimately obtained ( Figure 5 A):

[0142] Severe reconstruction (n=58): characterized by significant structural and functional impairment.

[0143] Functional relative compensation type (n=67): characterized by relative preservation of structural function.

[0144] R implementation: stats::kmeans() + cluster::silhouette() + ggplot2 for density distribution visualization, outputting Patient_Cluster_Label as the final classification label.

[0145] 7.5 Verification of Technical Effectiveness

[0146] Survival analysis confirmed that patients with more severe remodeling had significantly lower event-free survival than those with relatively compensated function (log-rank P = 0.0016). Figure 5 B) indicates that the cardiac remodeling classification based on CMR parameters constructed in this embodiment has clear accuracy and stratification value in predicting clinical outcomes.

[0147] The above results further demonstrate that the cardiac remodeling classification system not only has statistical stability but also verifiable predictive ability in clinical risk assessment. Therefore, it can serve as a prerequisite classification standard for subsequent screening of plasma protein biomarkers, construction of diagnostic models, and development of related testing products based on cardiac remodeling phenotypes.

[0148] 8. Screening of plasma proteins related to cardiac remodeling

[0149] Based on the aforementioned cardiac remodeling phenotypes, 1,309 differentially expressed proteins were selected as candidate variables. A least absolute contraction and selection operator (LASSO) regression model was used for feature screening to identify plasma proteins that could distinguish different remodeling phenotypes and possess potential prognostic value. Adjustments were made for age, sex, and NT-proBNP levels during the modeling process. Figure 6 AC). Through feature compression and cross-validation, five plasma proteins associated with stable cardiac remodeling were finally identified: MYL3, ANP, ALDH6A1, PPIA, and PKIA. Figure 7 These five proteins together constitute a core group of protein biomarkers closely related to the progression of cardiac remodeling and the risk of adverse outcomes.

[0150] 9. Prognostic follow-up and data collection

[0151] All patients were followed up for a median period of 17.1 months (interquartile range 8.5–26.3 months), and composite endpoint events including cardiovascular death, rehospitalization for heart failure, implantation of left ventricular assist device or heart transplantation were collected. A total of 28 prognostic events were recorded.

[0152] 10. Construction of a prognostic risk score based on five proteins

[0153] Based on the expression levels of five proteins (MYL3, ANP, ALDH6A1, PPIA, and PKIA) Figure 7 The protein risk score for each patient was calculated by combining a multivariate Cox regression model (adjusted for age, sex, and NT-proBNP).

[0154] Risk score calculation formula:

[0155] Risk Score = 0.4645 × ANP + 0.1174 × MYL3 - 0.3182 × PPIA - 0.1580× PKIA - 0.1516 × ALDH6A1

[0156] Wherein, Risk Score represents the risk score, ANP represents the expression level of ANP protein, MYL3 represents the expression level of MYL3 protein, PPIA represents the expression level of PPIA protein, PKIA represents the expression level of PKIA protein, and ALDH6A1 represents the expression level of ALDH6A1 protein.

[0157] Patients were divided into high-risk and low-risk groups based on the median risk score (cut-off = -0.469).

[0158] 11. Evaluation of prognostic efficacy

[0159] The association between protein risk score and prognostic events was assessed using a Cox proportional hazards model. Kaplan–Meier survival analysis showed that patients in the high-risk group had a significantly increased risk of the composite endpoint event compared to those in the low-risk group (HR = 3.95, 95% CI: 1.67–9.34, P = 0.001). Figure 8 The predictive power of the model was evaluated using Harrell's C index, which yielded a value of 0.79 (95% CI: 0.74–0.86), indicating that the model has good discriminative ability.

[0160] 12. External cohort prognostic validation based on cardiac remodeling classification and five-protein risk score

[0161] To further evaluate the stability and generalization ability of the prognostic risk stratification method based on multiple cardiac remodeling axes and five plasma protein biomarkers in an independent population, this embodiment validates the method in an external validation cohort. The external validation cohort included 91 patients with idiopathic dilated cardiomyopathy from Beijing Anzhen Hospital, and their clinical data collection, definition of follow-up endpoint events, and cardiac magnetic resonance imaging procedures were consistent with those of the discovery cohort.

[0162] During external validation, the key CMR parameters with large absolute loading values ​​and high contributions to the corresponding remodeling axis were first selected based on the cardiac remodeling axis loading coefficient matrices obtained from the discovery cohort through principal component analysis (PCA). These key parameters were standardized using the mean and standard deviation of the discovery cohort to eliminate the influence of dimensional differences. No model retraining or parameter re-estimation was performed during external validation.

[0163] Subsequently, based on the loading coefficients corresponding to each key CMR parameter in the discovery cohort, the standardized parameters were weighted and summed to reconstruct the scores of each subject in the external validation cohort on each cardiac remodeling axis. The first remodeling axis primarily reflects ventricular systolic function characteristics and is calculated based on LVEF, LVGCS, LVGRS, RVGCS, and RVGRS; the second remodeling axis primarily reflects ventricular structural remodeling characteristics and is calculated based on LVEDVI, LVESVI, LVMI, RVEDVI, RVESVI, and RVMI; the third remodeling axis primarily reflects right atrial structural and functional characteristics and is calculated based on RAVI, RAEF, and εs-RA; and the fourth remodeling axis primarily reflects myocardial histological characteristics and is calculated based on native T1 and ECV.

[0164] After obtaining multiple cardiac remodeling axis scores from patients in the external validation cohort, centroid mapping was used to map these patients to pre-defined cardiac remodeling phenotype categories in the discovery cohort, thus completing the external validation of the cardiac remodeling classification. Results showed that, consistent with the discovery cohort, patients with more severe remodeling phenotypes (n = 44) in the external validation cohort had significantly lower event-free survival than those with relatively compensated functional remodeling (n = 47) (log-rank P = 0.03). Figure 9 This suggests that the cardiac remodeling classification based on CMR parameters has good stability and consistent prognostic ability in independent populations.

[0165] Based on this, to verify the prognostic relevance and risk stratification ability of the five plasma protein biomarkers in an external cohort, this embodiment used enzyme-linked immunosorbent assay (ELISA) to quantitatively detect the plasma expression levels of ALDH6A1, MYL3, PPIA, PKIA, and ANP. The results showed that the above five proteins exhibited significant differential expression trends among patients with different cardiac remodeling phenotypes, and the direction of these changes was consistent with that of the cohort where the findings were observed. Figure 10 Further, based on the expression levels of the five proteins, a multivariate Cox regression model established in the discovery cohort was used to calculate the protein prognostic risk score for patients in the external cohort. Patients were then divided into high-risk and low-risk groups based on the median risk score. Kaplan-Meier survival analysis showed that the event-free survival rate of patients in the high-risk group was significantly lower than that of patients in the low-risk group (log-rank P = 0.042), indicating that the five-protein risk score also possesses good prognostic predictive ability and clinical applicability in an independent population. Figure 11 ).

[0166] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0167] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims, and the specification can be used to interpret the content of the claims.

Claims

1. The application of a reagent for detecting the expression level of protein biomarkers in the preparation of products for predicting the prognostic risk of patients with idiopathic dilated cardiomyopathy, wherein the protein biomarkers include ALDH6A1, PPIA, PKIA, MYL3, and ANP.

2. The application according to claim 1, characterized in that, The product is used to predict the risk of adverse cardiovascular events in patients with idiopathic dilated cardiomyopathy; The adverse cardiovascular events include one or more of the following: cardiovascular death, rehospitalization for heart failure, implantation of a left ventricular assist device, or heart transplantation.

3. The application according to claim 1 or 2, characterized in that, Compared with patients with idiopathic dilated cardiomyopathy who had a lower prognostic risk, patients with a higher prognostic risk had elevated levels of MYL3 and ANP, while the levels of ALDH6A1, PPIA, and PKIA were decreased.

4. The application according to claim 3, characterized in that, The sample was a plasma sample.

5. The application according to claim 1 or 2, characterized in that, The detection reagents are selected from one or more of the following methods: liquid chromatography, mass spectrometry, and immunoassay.

6. The application according to claim 5, characterized in that, Immunoassays include one or more of the following: Western blot analysis, radioimmunoassay, immunofluorescence assay, immunoprecipitation, immunodiffusion, electrochemiluminescence immunoassay, ELISA assay, and immunopolymerase chain reaction.

7. The application according to claim 5, characterized in that, The detection reagents are those applicable to liquid chromatography-tandem mass spectrometry (LC-MS / MS) or double antibody sandwich ELISA.

8. The application according to claim 5, characterized in that, The detection reagent contains specific antibodies or functional fragments thereof against MYL3, ANP, ALDH6A1, PPIA and PKIA.

9. The application according to claim 1 or 2, characterized in that, The products mentioned are reagent kits, chips, test strips, detection systems, or devices.

10. The application according to claim 9, characterized in that, The kit is a mass spectrometry labeled peptide detection kit or an enzyme-linked immunosorbent assay kit.

11. A method for constructing a prognostic risk scoring model for idiopathic dilated cardiomyopathy, characterized in that, The construction method includes using the expression levels of protein markers MYL3, ANP, ALDH6A1, PPIA, and PKIA.

12. The construction method according to claim 11, characterized in that, Based on the expression levels of the protein biomarkers used in patients with idiopathic dilated cardiomyopathy, the prognostic risk scoring model was constructed using Cox regression.

13. A prognostic risk scoring model for idiopathic dilated cardiomyopathy, characterized in that, The prognostic risk scoring model includes protein biomarkers such as MYL3, ANP, ALDH6A1, PPIA, and PKIA.

14. The prognostic risk scoring model according to claim 13, characterized in that, The prognostic risk scoring model calculates the risk score using the following formula: Risk Score = 0.4645 × ANP + 0.1174 × MYL3 - 0.3182 × PPIA - 0.1580 ×PKIA - 0.1516 × ALDH6A1 Wherein, Risk Score represents the risk score, ANP represents the expression level of ANP protein, MYL3 represents the expression level of MYL3 protein, PPIA represents the expression level of PPIA protein, PKIA represents the expression level of PKIA protein, and ALDH6A1 represents the expression level of ALDH6A1 protein.

15. The prognostic risk scoring model according to claim 14, characterized in that, Patients with a Risk Score ≥ -0.469 are classified as high-risk, while those with a Risk Score < -0.469 are classified as low-risk. High-risk patients have a significantly increased risk of cardiovascular death, heart failure readmission, left ventricular assist device implantation, or heart transplantation.

16. A predictive system for the prognostic risk of idiopathic dilated cardiomyopathy, characterized in that, The prediction system includes: The data acquisition module is used to acquire the expression levels of protein biomarkers in the samples of the subjects to be tested, including MYL3, ANP, ALDH6A1, PPIA and PKIA; The analysis module is used to provide the expression levels of plasma protein markers obtained by the data acquisition module as input data to the constructed prognostic risk scoring model; The output module is used to output the prognostic risk stratification results of the prognostic risk scoring model.

17. The prediction system according to claim 16, characterized in that, In the analysis module, the prognostic risk scoring model is modeled using Cox regression.

18. The prediction system according to claim 17, characterized in that, The prognostic risk scoring model is the prognostic risk scoring model obtained by the construction method described in claim 11 or 12, or the prognostic risk scoring model described in any one of claims 13-15.

19. A computer-readable storage medium, characterized in that, The system contains a computer program for performing the functions of the prediction system according to any one of claims 16-18.

20. A computer device, characterized in that, The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the functions of the prediction system according to any one of claims 16-18.

Images