Biomarker set, staging model and its application for chronic kidney disease

Abstract

This invention relates to the field of biomedical technology, and more particularly to a biomarker set, staging model, and their applications for chronic kidney disease. This invention utilizes metabolomics data analysis to obtain a biomarker set based on serum creatinine. Then, through methodological optimization, a staging model for chronic kidney disease is constructed. This staging model is also used for staging chronic kidney disease, and the results are accurate, specific, precise, and highly sensitive. It eliminates the false positive problem of traditional serum creatinine-based methods and removes the influence of gender factors. Staging chronic kidney disease based on this model is more consistent with the disease's progression characteristics.

Description

Technical Field

[0001] This invention relates to the field of biomedical technology, and in particular to a set of biomarkers, staging models and their applications for chronic kidney disease. Background Technology

[0002] Chronic kidney disease (CKD) is defined as persistent abnormalities in kidney structure or function (such as a glomerular filtration rate (GFR) <60 mL / min / 1.73 m³ / min). 2 Chronic kidney disease (CKD), characterized by kidney damage and progressive decline in renal function, is one of the most important and challenging kidney diseases. The prevalence of primary kidney disease continues to rise, and this trend is largely contributed by the increasing number of patients with metabolic disorders such as hypertension and diabetes, which contribute to secondary kidney injury. Approximately 40% of diabetic patients develop CKD, creating a significant healthcare burden. In addition to the nursing burden of kidney failure itself, diabetic nephropathy increases the risk of other comorbidities, particularly cardiovascular disease and infections. Despite advances in understanding the epidemiology of CKD, current treatment strategies remain primarily focused on symptom management and slowing disease progression. Long-term progression of CKD can lead to renal insufficiency, eventually progressing to end-stage renal disease, requiring dialysis or transplantation. This not only increases the burden on healthcare providers but also poses a significant financial challenge to patients' families. Therefore, it is essential to understand the prognosis and staging of patients with chronic kidney disease, especially since patients in the later stages (CKD4-CKD5) have a poorer prognosis, making early diagnosis and later treatment crucial.

[0003] Currently, two staging systems for chronic kidney disease are commonly used in clinical practice: one is based on serum creatinine (CKD-EPI creatinine equation), and the other is based on serum cystatin (CysC, CKD-EPI CysC equation). Both indirectly calculate glomerular filtration rate by detecting substrate concentration in peripheral blood, without invasive procedures such as kidney biopsy, making them suitable for routine screening and monitoring. However, these two staging systems rely too heavily on a single indicator, which can easily lead to false staging results, and gender factors have a significant impact on the calculation results.

[0004] Therefore, it is particularly necessary to develop a new set of multi-indicator and highly accurate chronic kidney disease staging biomarkers and staging assessment models. Summary of the Invention

[0005] In view of this, the technical problem to be solved by the present invention is to provide a set of biomarkers, a staging model and their applications for chronic kidney disease.

[0006] This invention provides a set of biomarkers for the prognosis and / or staging of chronic kidney disease, including: age, CD3... + CD8 + .MMP low # Hemoglobin, serum creatinine, and serum potassium, excluding the sex coefficient.

[0007] Furthermore, the CD3 + CD8 + .MMP low The unit of # is cells / μL, the unit of hemoglobin is g / L, the unit of serum creatinine is μmol / L, and the unit of serum potassium is mmol / L.

[0008] In this invention, the CD3 + CD8 + .MMP low # Absolute count of low membrane potential inhibitory T cells (cells / mL) or CD3 + CD8 + Absolute count of cells with low membrane potential (cells / mL).

[0009] This invention provides a method for screening a set of markers, comprising the following steps:

[0010] Step 1: Analyze the correlation between the biomarkers to be screened and the glomerular filtration rate, and select biomarkers with a correlation coefficient ρ < -0.14 or ρ > 0.1 as the candidate set 1;

[0011] Step 2: Through random forest model analysis, the markers in candidate set 1 are sorted according to their contribution to the G1~G5 groupings, and the marker set described in this invention is obtained by screening.

[0012] Furthermore, the parameters of the random forest model are set as follows: when n=5, the cross-validation error is the lowest, and the average reduction in precision is >37.

[0013] In this invention, the p-value is always less than 0.05 when the correlation coefficient ρ < -0.14, and is further limited to p being less than 0.01; the p-value is always less than 0.05 when the correlation coefficient ρ > 0.14, and is further limited to p being less than 0.01.

[0014] This invention provides the application of the biomarker set described above and / or the biomarker set obtained by the screening method described above in constructing a staging model for chronic kidney disease.

[0015] This invention provides a staging model for chronic kidney disease, the formula for which is:

[0016] Formula I;

[0017] or , Formula II;

[0018] In Formula I, ICKDSI is the immune nephropathy index; CD3 + CD8 + .MMP low #CD3 concentration per milliliter of sample + CD8 + The absolute count of T cells with low membrane potential; HB is the hemoglobin content in grams per liter of sample; Age is the patient age corresponding to the sample; K is the serum potassium content in mmol per liter of sample; Scr is the serum creatinine content in μmol per liter of sample.

[0019] In formula II, The values ​​are the min-max standardized values ​​of the patients' ages corresponding to the sample. The absolute count of low membrane potential inhibitory T cells per milliliter of sample is the min-max normalized value. The value is the min-max normalized value of the hemoglobin content in grams per liter of sample. The value is the serum creatinine content in μmol per liter of sample, standardized by min-max. The value is the blood potassium content in mmol per liter of sample, standardized by min-max.

[0020] In the staging model described in this invention, the criteria for staging chronic kidney disease using the staging model are as follows:

[0021] In Formula I, an ICKDSI ≥ 163.23 indicates stage 1 chronic kidney disease (CKD).

[0022] A score of 103.93 < ICKDSI < 163.22 indicates stage 2 chronic kidney disease (CKD).

[0023] 86.85≤ICKDSI≤103.92 indicates chronic kidney disease (CKD) stage 3a.

[0024] A score of 57.41 ≤ ICKDSI < 86.84 indicates chronic kidney disease (CKD) stage 3b.

[0025] A score of 32.43 ≤ ICKDSI < 57.41 indicates stage 4 chronic kidney disease (CKD).

[0026] An ICKDSI score of <32.42 indicates stage 5 chronic kidney disease (CKD).

[0027] In Formula II, PC1 ≤ -0.092 indicates stage 1 chronic kidney disease (CKD).

[0028] -0.092>PC1≤0.054, indicating stage 2 chronic kidney disease (CKD);

[0029] 0.054 > PC1 < 0.141 indicates chronic kidney disease (CKD) stage 3a.

[0030] 0.141 ≥ PC1 < 0.171 indicates chronic kidney disease (CKD) stage 3b.

[0031] 0.171 ≥ PC1 < 0.306 indicates stage 4 chronic kidney disease (CKD).

[0032] PC1 ≥ 0.3057 indicates CKD5 stage 5 chronic kidney disease.

[0033] The present invention provides a detection reagent that uses a set of markers as described in the present invention and / or a set of markers obtained by screening as described in the present invention as the detection target.

[0034] This invention provides a detection product comprising the detection reagents as described herein, as well as acceptable auxiliaries, carriers, or devices.

[0035] This invention provides the application of the aforementioned biomarker set, the biomarker set obtained by the aforementioned screening method, the aforementioned staging model, the aforementioned detection reagent, and / or the aforementioned detection product in the preparation of kits for the prognosis and / or staging of chronic kidney disease.

[0036] This invention provides the application of the protein combination, the set of biomarkers obtained by the screening method, the prediction model, the detection reagent, and / or the detection product in constructing a system for the prognosis and / or staging of chronic kidney disease.

[0037] This invention provides a system for the prognosis and / or staging of chronic kidney disease, comprising: a data collection module, a data analysis module, and a result output module;

[0038] The data collection module acquires the age of the subjects from the biomarker set described in this invention, and the CD3 levels in the subjects' blood samples. + CD8 + .MMP low # Data on hemoglobin, serum creatinine, and serum potassium levels;

[0039] The data analysis module: based on the subject's age and the CD3 levels in the subject's blood sample... + CD8 + .MMP low # Data on hemoglobin, serum creatinine, and serum potassium levels are used to calculate the immune nephropathy index and / or PCI using the staging model described in this invention;

[0040] The result output module stages chronic kidney disease based on the magnitude of the immune nephropathy index.

[0041] This invention provides the application of the aforementioned biomarker set, the biomarker set obtained by the aforementioned screening method, the aforementioned staging model, the aforementioned detection reagent, the aforementioned detection product, and / or the aforementioned system in the preparation of equipment for the prognosis and / or staging of chronic kidney disease.

[0042] A method for staging chronic kidney disease for non-diagnostic purposes, which involves staging chronic kidney disease using the staging model and / or the system described in this invention;

[0043] Furthermore, the method for staging chronic kidney disease according to the present invention includes the following steps:

[0044] Step 1: Obtain the subject's age from the biomarker set described in this invention, and the CD3 level in the subject's blood sample. + CD8 + .MMP low # Data on hemoglobin, serum creatinine, and serum potassium levels;

[0045] Step 2: Stage chronic kidney disease using the staging model and / or the system described in this invention.

[0046] This invention utilizes metabolomics data analysis to obtain a set of biomarkers based on serum creatinine. Then, through methodological optimization, a staging model for chronic kidney disease (CKD) is constructed. This staging model addresses the problem that traditional glomerular filtration rate (GFR) staging based on serum creatinine relies too heavily on a single indicator, easily leading to false staging results. Simultaneously, by using multiple indicators in combination, it compensates for the inadequacy of single-indicator clustering and eliminates the influence of gender factors in the calculation process (no gender coefficient is required). The staging model constructed in this invention has a correlation of over 80% with the original staging, and staging based on this model better reflects the characteristics of disease progression. The immune nephropathy index (INI) or PCI values ​​obtained through the staging model have greater variability, stronger correlation, and similar areas under the curve, while effectively distinguishing the differences in data related to whether or not there is an immune response, objectively differentiating different stages of CKD, which is superior to using a single serum creatinine indicator.

[0047] This invention utilizes metabolomics data analysis to obtain a set of biomarkers based on serum creatinine. Then, through methodological optimization, a staging model for chronic kidney disease was constructed. This staging model is also used for the staging of chronic liver disease, and the results are accurate, specific, precise, and sensitive. It eliminates the false positive problem of the original serum creatinine-based indicators and removes the influence of gender factors. The staging of chronic kidney disease based on this model is more consistent with the development characteristics of the disease. Attached Figure Description

[0048] Figure 1 The correlation between serum creatinine and glomerular filtration rate;

[0049] Figure 2 The following diagram illustrates the screening of immunological markers using flow cytometry: A) Whole blood event overview—showing forward-lateral scattering to delineate the leukocyte population; B) Delineation of lymphocytes; C) Exclusion of adhesion cells; D) CD3. + T cell screening—showing the percentage of total T cells; E shows CD4 vs CD8 subset breakdown—showing the distribution of helper-killer T cells; F shows CD4... + PD-1 + Exhaustion subsets—indicating the expression intensity of immune checkpoints; G represents CD8. + PD-1 + Exhausted subsets—indicating cytotoxic T cell function suppression; H represents the MitoDye functional axis—an assessment of T cell mitochondrial activity;

[0050] Figure 3 Demonstration of the methodology flowchart;

[0051] Figure 4 This indicates immunological and serological indicators related to glomerular filtration rate (GFR) values;

[0052] Figure 5 The indicators related to glomerular filtration rate values ​​were ranked according to their contribution to the G1-G5 groups;

[0053] Figure 6 This paper presents a differential analysis of five key indicators selected through random forest analysis across different time periods; where A represents score (SCR); B represents age; C represents HB; D represents K; and E represents CD3. + CD8 + .MMP low #;

[0054] Figure 7 The ROC curves of expression profiles of five main indicators, PC1 to PC5, at different stages are shown.

[0055] Figure 8 The correlation between Formula 1, Formula 2, and Formula PC1 and commonly used clinical indicators such as glomerular filtration rate and serum creatinine is shown.

[0056] Figure 9 Indicates the staging index of immune-related chronic kidney disease;

[0057] Figure 10 This section compares the Youden indices of Formula 1 and Formula 2 across three disease sets and in the case of diabetic nephropathy alone; where A represents the difference in Youden indices between Formula 1 and Formula 2 across the three disease sets (lgA nephropathy, membranous nephropathy, and diabetic nephropathy) (V). Wilcoxon=3.00, p=0.08, CI 95 % [-0.96, -0.23]); B. The difference in Youden's index between Formula 1 and Formula 2 in diabetic nephropathy as a single disease (V Wilcoxon =5.00, p=1.00, CI 95 %[-0.68, -0.68]);

[0058] Figure 11 This study validated the ICKDSI formulas 1, 2, and 3 and the serum creatinine group in a population with chronic kidney disease and diabetes (n=219). In this study, A represents the Youden index; B represents sensitivity; C represents precision; D represents the area under the curve; E represents specificity; and F represents accuracy. Detailed Implementation

[0059] This invention provides a set of biomarkers, a staging model, and their applications for chronic kidney disease. Those skilled in the art can refer to this document and appropriately modify the process parameters to achieve the desired results. It is particularly important to note that all similar substitutions and modifications are obvious to those skilled in the art and are considered to be included in this invention. The methods and applications of this invention have been described through preferred embodiments. Those skilled in the art can clearly modify or appropriately change and combine the methods and applications described herein without departing from the content, spirit, and scope of this invention to implement and apply the technology of this invention.

[0060] Terminology Explanation:

[0061] AUC: Area under the curve, a core indicator for quantifying the discriminative power of diagnostic tests.

[0062] CD3 + CD8 + .MMP low #: Absolute count of low membrane potential inhibitory T cells (cells / mL).

[0063] CKD: Chronic kidney disease.

[0064] CKD1~CKD5: Stages of chronic kidney disease, with the condition progressively worsening from CKD1 to CKD5.

[0065] CKD-EPI: The preferred formula for estimating glomerular filtration rate (eGFR), developed by the Chronic Kidney Disease Epidemiology Collaborative Research Group, is more accurate than the traditional MDRD formula.

[0066] CysC: Serum cystatin is a precise and sensitive indicator for assessing kidney function. It is not affected by factors such as muscle mass or diet and can reflect subtle changes in glomerular filtration function earlier.

[0067] DN: Diabetic Nephropathy.

[0068] EDTA: Chelating agents are commonly used as blood anticoagulants.

[0069] eGFR: Glomerular filtration rate (GFR or eGFR) is a measure of glomerular filtration rate, expressed in mL / min / 1.73m³. 2 .

[0070] HB: Hemoglobin, a core indicator for assessing anemia in the human body. It is mainly responsible for transporting oxygen in the blood. Abnormal levels of HB can directly affect the body's oxygen supply and recovery and risk assessment in situations such as postoperative care and kidney disease.

[0071] HDL / LDL: High-density lipoprotein / Low-density lipoprotein.

[0072] ICKDSI: Immune-related chronic kidney disease staging index (ICKDSI, hereinafter referred to as "immuno-nephropathy index").

[0073] IgAN: IgA nephropathy: a common type of kidney disease characterized by abnormal deposition of immunoglobulin A (IgA) in the glomerular mesangial area.

[0074] K: Serum potassium, refers to the concentration of potassium ions in the blood, with a normal range of 3.5~5.5 mmol / L.

[0075] min-max: A data standardization method that makes the overall range of data between 0 and 1.

[0076] MitoDye: A mitochondrial probe, a small molecule fluorescent dye that can specifically bind to the inner mitochondrial membrane based on the mitochondrial membrane potential difference and form a covalent bond.

[0077] MMP low #: Absolute count of cells with low mitochondrial membrane potential.

[0078] MMP low %: Percentage of mitochondrial low membrane potential.

[0079] MN: Membranous Nephropathy.

[0080] PCA: Principal Component Analysis, a core statistical method commonly used for data dimensionality reduction and feature extraction.

[0081] ROC: Subject observation curve.

[0082] Scr: Serum creatinine. Creatinine is the metabolic end product of creatine in muscle tissue. It is mainly excreted through glomerular filtration and is hardly reabsorbed by the renal tubules. Its blood concentration is closely related to muscle mass and kidney function. For men, it is 53~106 μmol / L (0.6~1.2 mg / dL), and for women, it is 44~97 μmol / L (0.5~1.1 mg / dL).

[0083] ρ value: correlation coefficient, using the Spearman method.

[0084] Advantages of this invention:

[0085] 1. As can be seen from Formula 2 (Formula I of this invention), ICKDSI calculation does not require data standardization. Compared with the calculation of glomerular filtration rate value commonly used in clinical practice, which does not involve cumbersome calculations such as gender coefficients (e.g., racial coefficients, coefficient κ, etc.), the test indicators can be directly substituted into the formula to calculate the ICKDSI value. The relevant staging results can be automatically output, which is more convenient and faster.

[0086] 2. By combining immunological and serological multi-indicator analysis, five important indicators were selected from 44 clinical test indicators. These indicators have better stability and a lower false positive rate than single indicators. Furthermore, they compensate for the fact that the original indicators were calculated based on a single blood indicator, which ignored the important factor that the immune system participates in chronic kidney disease and affects disease progression and immune recovery through inflammation.

[0087] 3. Accuracy and consistency of ICKDSI values ​​in staging: (1) Accuracy. The staging characteristics of the data were verified by grouping diabetic and non-diabetic patients, and the results were consistent; (2) Consistency. We verified the results using a test dataset (n=340) and an external dataset (n=346), and the results were consistent.

[0088] The test materials used in this invention are all common commercially available products. The invention is further illustrated below with reference to embodiments:

[0089] Example 1: Traditional analysis method and the analysis method of the present invention

[0090] I. Calculation of traditional glomerular filtration rate indicators.

[0091] The glomerular filtration rate (eGFR) value was calculated using the CKD-EPI (2021 recommendation) formula, targeting an eGFR >60 mL / min / 1.73 mcg. 2 This is a commonly used clinical calculation method for patients. The main variables include: serum creatinine (Scr, μmol / L or mg / dL, hereinafter referred to as "serum creatinine or Scr"), age, and sex. The calculation formula is:

[0092] Glomerular filtration rate = 141 × min(Scr / κ,1) α ×max(Scr / / κ,1) -1.209 ×0.993 年龄 × Gender coefficient, Formula A;

[0093] Where, Scr (unit: mg / dL) = Scr (μmol / L) / 88.4; κ = 0.7 for females, κ = 0.9 for males; α = 0.7 for females, α = 0.9 for males; sex coefficient (女性) =1.018, gender coefficient (男性) =1. Analysis was conducted based on the glomerular filtration rate (GFR) staging system, referencing chronic kidney disease (CKD) groupings, and using the standardized quantitative values ​​(data rounded to two decimal places). The calculated GFR values ​​and staging methods are as follows: G1 (CKD1): GFR > 90; G2 (CKD2): GFR 60–90; G3a (CKD3a): GFR 45–60; G3b (CKD3b): GFR 30–45; G4 (CKD4): GFR 15–30; G5 (CKD5): GFR < 15.

[0094] Second, both the formula and the calculated values ​​show that the calculated glomerular filtration rate is positively correlated with the serum creatinine level.

[0095] Figure 1 The correlation between serum creatinine and glomerular filtration rate is shown (from experimental data, S=1.28e+07, P=2.75e-185). The x-axis represents the glomerular filtration rate (clinically calculated value), and the y-axis represents the serum creatinine value (clinically measured serum value). The correlation analysis used Spearman correlation analysis, with a ρ value of -0.96 (ρ = -1 indicates an absolute negative correlation).

[0096] Third, the current method of staging based on eGFR calculation has been reported to have certain limitations regarding glomerular filtration rate calculated based on serum creatinine.

[0097] A recent report by Ávila et al. points out that creatinine levels are limited because they are influenced by factors such as advanced age, physical activity, protein-rich diets, male sex, medications, and ethnicity. On the one hand, certain medications, such as cimetidine, trimethoprim, and abecilibil, inhibit renal tubular secretion of Cr through competitive secretion pathways, thereby increasing serum creatinine levels and leading to an inaccurate estimate (decreased) of GFR. On the other hand, Cr levels vary with muscle mass, dietary protein intake, and creatine supplementation. Levels tend to decrease with age in women and among different individuals. Nishiwaki et al. also note that in real-world clinical settings, the frequency and intervals of glomerular filtration rate (GFR) measurements can vary significantly between individuals. Because such discrepancies may reflect systemic changes in the risk of adverse patient outcomes, estimates of GFR variability based on clinical measurements may be biased and difficult to implement.

[0098] IV. This invention adds immunological and serological detection indicators to the original serum creatinine.

[0099] Based on this, this invention conducted a study on common chronic kidney disease types including IgA nephropathy (IgAN), membranous nephropathy (MN), and diabetic nephropathy (DN), and explored a set of peripheral biomarkers related to chronic kidney disease staging in common indicators of these diseases, mainly including immunological indicators and serum indicators.

[0100] 1. Obtaining immunological indicators

[0101] Immunological markers were obtained using flow cytometry. The flow cytometry detection procedure was as follows: Collect 2-5 mL of sample and immediately mix gently to prevent clotting. Gently mix, verify the absence of clots and hemolysis, confirm the accuracy of basic information, and then proceed to subsequent steps.

[0102] Sample pretreatment:

[0103] (1) 100 μL of peripheral blood specimen containing EDTA anticoagulant was treated with erythrocyte lysis buffer and then incubated with mixed antibodies at room temperature in the dark for 15 min;

[0104] (2) Add 2 mL of hemolysin to destroy red blood cells.

[0105] (3) Centrifuge the sample at 300×g for 5 min. Discard the supernatant, resuspend the precipitate in 200 μL PBS, transfer it into a tube containing MitoDye, and incubate at 37°C in the dark for 30 min.

[0106] (4) Finally, transfer it to the flow cytometer and then test it on the machine. Data is collected by using the applicable template and the gate.

[0107] The reagents used for detection were CD3 FITC, CD45 PerCP-Cy5.5, CD4 PE-Cy7, CD8 APC-Cy7, PD-1 PE, and MitoDye fluorescent probes (Ubiquitous Biotech, Ningbo, China). The set of these probes (CD3 / CD4 / CD8 / CD45 / PD-1 / MitoDye combination) was used to analyze "PD-1⁺ T cells with mitochondrial dysfunction", i.e., suppressor T cells.

[0108] The instrument used for detection was a NovoCyte flow cytometer (Agilent Technologies, Santa Clara, CA, USA). Flow cytometry gating strategy reference. Figure 2 .

[0109] Absolute count indicators were determined by counting labeled immune cells using an instrumental volumetric method. At least 8000 events were collected. Final analysis and graphical output were performed using NovoExpress software (Agilent Technology, USA), followed by direct output via "Human Lymphocyte Mitochondrial Function Assessment Data Analysis Software," showing the percentage of each cell subset relative to lymphocytes. The absolute count indicator for mitochondrial low membrane potential cells (MMP) was also included. low The cell count of # is the absolute cell count of the next higher phylum multiplied by MMP. low % Calculated.

[0110] The results are as follows Figure 2 As shown, where, Figure 2 AG in the figure represents target cell analysis. Figure 2 H in the image represents the target cells (with CD3). + (Taking T cells as an example) Mitochondrial markers (M7-1 is CD3) + .MMP low The mitochondrial index (including mitochondrial quality index and mitochondrial low membrane potential percentage index) is the median fluorescence intensity value of all fluorescence signals.

[0111] 2. Screening of serum markers

[0112] Valid data for routine clinical serological markers include:

[0113] (1) The proportion and count of white blood cells, lymphocytes and their subsets, and the corresponding mitochondrial mass and percentage of low-Mitochondrial membrane potential (MMP) of the cells. low %);

[0114] (2) Biochemical indicators, such as hemoglobin (HB), serum creatinine (Scr), uric acid (UA), albumin (Alb), triglycerides (TG), total cholesterol (TCH), high-density lipoprotein (HDL), low-density lipoprotein (LDL), etc.

[0115] (3) Ion indicators, such as potassium, calcium, phosphorus and other indicators.

[0116] V. Calculation of Immune Nephropathy Index

[0117] The ICKDS1 value can be calculated by directly inputting serum creatinine and the selected immunological and serological test indicators into formula B, which has been optimized by methodology.

[0118] Formula B;

[0119] The corresponding classification of the immune nephropathy index values ​​calculated using this formula is as follows: CKD1: ICKD1 ≥ 163.23; CKD2: 103.93 < ICKD1 < 163.22; CKD3a: 86.85 ≤ ICKD1 ≤ 103.92; CKD3b: 57.41 ≤ ICKD1 < 86.84; CKD4: 32.43 ≤ ICKD1 < 57.41; CKD5: ICKD1 < 32.42.

[0120] Example 2: Screening of immunological and serological detection indicators and methodological optimization of the immune nephropathy index

[0121] The methodological optimization process for screening immunological and serological detection indicators and the immune nephropathy index is as follows: Figure 3 As shown, the detailed steps are as follows:

[0122] 340 CDK patients were used as the test set, and 346 external validation patients were used for validation using established staging. The test set patients were categorized into G1–G5 (CKD1, CKD2, CKD3a, CKD3b, CKD4, CKD5) based on their glomerular filtration rate (eGFR) staging, and correlation analysis was performed. First, candidate indicators strongly correlated with glomerular filtration rate were screened through significance and correlation analysis. Then, glomerular filtration rate correlation analysis was conducted based on p... Spearman After ranking the correlation coefficients of 44 immune and serological indicators, 31 indicators with correlation coefficients ρ < -0.14 or ρ > 0.14 and p < 0.01 were selected. Figure 4 Then, random forest analysis is used to rank these indicators according to their contribution importance to the G1~G5 groups, i.e., the mean decline accuracy from high to low. The lowest cross-validation error is obtained when n=5, at which point the average reduction in accuracy is >37. Figure 4 The five best indicators strongly correlated with glomerular filtration rate were: serum creatinine (Scr), age, hemoglobin (HB), and CD3. + CD8 + .MMP low # Serum potassium (K) level ( Figure 5 ); and the differences in the five key indicators screened through random forest analysis and validation across different stages are shown in the following figures. Figure 6 As shown, these five indicators exhibit differential expression across different stages and can be used for the G1-G5 staging of chronic kidney disease.

[0123] Then, principal component analysis was used to analyze the five main indicators (serum creatinine (Scr), age, hemoglobin (HB), CD3) + CD8 + .MMP low # The serum potassium (K) index was normalized using the min-max method and dimensionality reduced to obtain five principal components (PC1~PC5) (Table 1). By comparing the area under the receiver operating characteristic (ROC) curves of these five principal components, the expression profile of PC1 was finally confirmed, that is, PC1 is a strongly correlated index of glomerular filtration rate and can be used for G1~G5 staging (Table 1, ...). Figure 7 The formula for calculating PC1 is as follows:

[0124] PC1 = 0.3372×x1 - 0.376×x2 - 0.5252×x3 + 0.5064×x4 + 0.461×x5, Formula C;

[0125] Where x1~x5 are the age and CD3 values ​​after min-max normalization, respectively.+ CD8 + .MMP low # Hemoglobin (HB, g / L), serum creatinine (Scr, μmol / L), serum potassium (K, mmol / L); The staging criteria based on PCI values ​​are as follows: PCI value ≤ -0.092, chronic kidney disease (CKD) stage 1; -0.092 > PC1 ≤ 0.054, chronic kidney disease (CKD) stage 2; 0.054 > PC1 < 0.141, chronic kidney disease (CKD) stage 3a; 0.141 ≥ PC1 < 0.171, chronic kidney disease (CKD) stage 3b; 0.171 ≥ PC1 < 0.306, chronic kidney disease (CKD) stage 4; PC1 ≥ 0.3057, chronic kidney disease (CKD) stage 5.

[0126] The coefficients calculated by the formula are shown in Tables 2 and 3. In Table 2, C1 to C5 are principal components 1 to 5, respectively.

[0127] Table 1. Principal Component Analysis Data: Principal Component Proportions and Component Characteristics (Loading Matrix)

[0128]

[0129] Table 2. Principal Component Composition

[0130]

[0131] Table 3. Principal Component Calculation Coefficients

[0132]

[0133] Note: Coefficient 1 = C1 / The calculation for coefficients 2 to 5 is the same as here.

[0134] Because principal component analysis is relatively complex (requiring the detection index values ​​to undergo min-max standardized data transformation before calculation), it is simplified based on division operations (i.e., based on the correlation with glomerular filtration rate, positive correlation is used as the numerator and negative correlation as the denominator). This is achieved by performing common logarithmic operations (lg(index value)) and square root operations on the index values. Squaring (indicators) nAfter reducing computational bias, a machine learning-based formula for calculating the Immune-related Chronic Kidney Disease Staging Index (ICKDSI) was developed. The ICKDSI value can be calculated by directly substituting the indicators into the formula. By processing four key indicators from two sets of pattern detection data combined with age (a total of five indicators), the following three ICKDSI calculation formulas were obtained:

[0135]

[0136] Based on the calculation of glomerular filtration rate using serum creatinine, a methodology for developing the "immuno-nephropathy index" was developed. Principal component analysis (PCA) of a network was used to calculate the new index, the immune nephropathy index. The correlation between the index and commonly used clinical indicators such as glomerular filtration rate and serum creatinine was analyzed using values ​​obtained through different methods. A p-value > 0.6 or < -0.6, with p < 0.05, was considered a correlation. Principal component 1 (PC1, obtained through principal component dimensionality reduction analysis) showed a good correlation with glomerular filtration rate (ρ = -0.768, p < 0.001), but it still lagged significantly behind other correlation coefficients. For example, in terms of correlation, formulas 1 and 2 (formula B) showed a lower correlation than other indicators (…). Figure 8 Mesoglomerular filtration rate, serum creatinine, and PC1 levels are strongly correlated with glomerular filtration rate (ρ). 公式1 =0.972, ρ 公式2 =0.972) and exceeds the correlation between serum creatinine and glomerular filtration rate (ρ). 血肌酐 =0.958, Figure 8 ).

[0137] The immune nephropathy index values ​​obtained from principal component analysis and three different formulas were compared. The immune nephropathy index values ​​obtained using formulas (e.g., Formula 1, Formula 2) showed greater variability, stronger correlation, and similar areas under the curve (Table 4) compared to the combined analysis-generated immune nephropathy index values ​​(PC1). This formula effectively distinguishes the differences in data related to whether or not there is an immune response, and objectively differentiates different stages of chronic kidney disease. Figure 9 (Tables 4 and 5), and it is superior to the single serum creatinine indicator (Tables 4 and 5). Furthermore, correlation analysis also shows that the correlation coefficients ρ between Formula 1 and Formula 2 and glomerular filtration rate are... (Spearman) The correlation coefficient between serum creatinine and glomerular filtration rate is higher than that between ρ and p. 公式1 =0.972, ρ 公式2 =0.974, all > ρ 血肌酐=0.958). Combining the two formulas to calculate the relevant Youden index and sensitivity values, and comparing them in the staging of three diseases (IgA nephropathy, membranous nephropathy, and diabetic nephropathy) and the staging of diabetic nephropathy alone, both Formula 1 and Formula 2 (Tables 4 and 5) were superior to the serum creatinine index ( Figure 10 ).

[0138] Table 4. Analytical parameters for serum creatinine, Formula 1, Formula 2, and PC1 between stages in the three diseases.

[0139]

[0140]

[0141] Table 5. Analytical parameters for serum creatinine, Formula 1, Formula 2, and PC1 between stages in the three diseases.

[0142]

[0143]

[0144] Considering the impact of diabetes on mitochondrial indicators, patients with chronic kidney disease and diabetes (n=219) and patients with diabetic nephropathy were combined into group 1 (diabetic nephropathy group), and patients with chronic kidney disease without diabetes were combined into group 2 (non-diabetic nephropathy group). Formulas 1 and 2 were then validated separately. In the diabetic nephropathy group, the Youden index (raw values ​​as shown in Table 6), sensitivity (raw values ​​as shown in Table 7), precision (raw values ​​as shown in Table 8), area under the curve (raw values ​​as shown in Table 9), specificity (raw values ​​as shown in Table 10), and accuracy (raw values ​​as shown in Table 11) of the immune nephropathy index were calculated by comparing the two formulas between adjacent stages (e.g., CKD1 vs. CKD2, CKD2 vs. CKD3a, CKD3a vs. CKD3b, etc.). Figure 11 The results showed that Formula 1 had better sensitivity than Formula 2 in the diabetic nephropathy group (CKD1 vs. CKD2: 90.91 vs. 90.91; CKD2 vs. CKD3a: 90.32 vs. 83.87; CKD3a vs. CKD3b: 74.47 vs. 70.21; CKD3b vs. CKD4: 94.29 vs. 94.29; CKD4 vs. CKD5: 87.50 vs. 95.83, Tables 6-11), with no significant advantage in accuracy, precision, or Youden's index. Similarly, in the non-diabetic nephropathy group, Formula 2 had better sensitivity than Formula 1 for CKD1 vs. CKD2 time periods (97.59 vs. 94.87), and it was still impossible to determine which formula was more advantageous from these results.

[0145] Table 6. Specific Values ​​of the Yoden Index

[0146]

[0147] Table 7. Specific values ​​of sensitivity (%)

[0148]

[0149] Table 8. Specific values ​​for accuracy (%)

[0150]

[0151] Table 9. Specific values ​​of area under the curve (%)

[0152]

[0153] Table 10. Specificity (%) values

[0154]

[0155] Table 11. Accuracy (%) Specific Values

[0156]

[0157] The diabetic nephropathy group and the non-diabetic nephropathy group were staged using critical values ​​to verify their consistency with glomerular filtration rate. The staging range of Formula 1 with critical values ​​was CKD1>136.62>CKD2>66.80>CKD3a>36.80>CKD3b>22.63>CKD4>8.34>CKD5; similarly, the staging range of Formula 2 with critical values ​​was CKD1>163.23>CKD2>103.93>CKD3a>86.85>CKD3b>57.41>CKD4>32.43>CKD5. For details on the endpoint staging, please refer to Example 1. In the diabetic nephropathy group, 26 and 25 cases, respectively, showed discrepancies between the parameters of Formula 1 and Formula 2 and the glomerular filtration rate (GFR) staging (corresponding concordance rates of 88.128%). In the non-diabetic nephropathy group, the number of cases with discrepancies between Formula 1 and Formula 2 were 20 and 14, respectively, with concordance rates of 83.471% and 88.430% (Table 12). Subsequently, we further validated the consistency between Formula 2 and GFR staging by adding external data on chronic kidney disease (n=346). A total of 66 cases were found to have discrepancies in staging (consistency rate of 80.925%). After removing the classification gray area (or the range prone to bias) within the threshold ±5, 48 out of the remaining 294 cases did not have discrepancies, with a concordance rate of 83.674% (Table 5). Based on the above, Formula 2 (Formula B) was ultimately selected as the method for obtaining the immune nephropathy index.

[0158] Table 12. Error rate and consistency of analyzing Formula 1, Formula 2 and glomerular filtration rate staging in Group 1, Group 2 and external validation group.

[0159]

[0160] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A set of biomarkers for the prognosis and / or staging of chronic kidney disease, characterized in that, include: Age, CD3 + CD8 + .MMP low #, hemoglobin, serum creatinine, and serum potassium, excluding the sex coefficient; CD3 + CD8 + .MMP low #CD3 concentration per milliliter of sample + CD8 + The absolute count of T cells with low membrane potential.

2. The method for screening a set of markers as described in claim 1, characterized in that, Includes the following steps: Step 1: Analyze the correlation between the biomarkers to be screened and the glomerular filtration rate, and select biomarkers with a correlation coefficient ρ < -0.14 or ρ > 0.14 as the candidate set 1; Step 2: Using a random forest model, the markers in the candidate set 1 are sorted according to their contribution to the G1~G5 groups, and the marker set described in claim 1 is obtained by screening.

3. The application of the biomarker set as described in claim 1 or the biomarker set obtained by the screening method as described in claim 2 in the preparation of products that construct staging models for chronic kidney disease.

4. The application according to claim 3, characterized in that, The phased model is as follows: Formula I; or , Formula II; In Formula I, ICKDSI is the immune nephropathy index; CD3 + CD8 + .MMP low #CD3 concentration per milliliter of sample + CD8 + The absolute count of T cells with low membrane potential; HB is the hemoglobin content in grams per liter of sample; Age is the patient age corresponding to the sample; K is the serum potassium content in mmol per liter of sample; Scr is the serum creatinine content in μmol per liter of sample. In formula II, The values ​​are the min-max standardized values ​​of the patients' ages corresponding to the sample. The absolute count of low membrane potential inhibitory T cells per milliliter of sample is the min-max normalized value. The value is the min-max normalized value of the hemoglobin content in grams per liter of sample. The value is the serum creatinine content in μmol per liter of sample, standardized by min-max. The value is the blood potassium content in mmol per liter of sample, standardized by min-max.

5. The application according to claim 4, characterized in that, The criteria for staging chronic kidney disease using the aforementioned staging model are as follows: In Formula I, an ICKDSI ≥ 163.23 indicates stage 1 chronic kidney disease (CKD). A score of 103.93 < ICKDSI < 163.22 indicates stage 2 chronic kidney disease (CKD). 86.85≤ICKDSI≤103.92 indicates chronic kidney disease (CKD) stage 3a. A score of 57.41 ≤ ICKDSI < 86.84 indicates chronic kidney disease (CKD) stage 3b. A score of 32.43 ≤ ICKDSI < 57.41 indicates stage 4 chronic kidney disease (CKD). An ICKDSI score of <32.42 indicates stage 5 chronic kidney disease (CKD). In Formula II, PC1 ≤ -0.092 indicates stage 1 chronic kidney disease (CKD). -0.092>PC1≤0.054, indicating stage 2 chronic kidney disease (CKD); 0.054 > PC1 < 0.141 indicates chronic kidney disease (CKD) stage 3a. 0.141 ≥ PC1 < 0.171 indicates chronic kidney disease (CKD) stage 3b. 0.171 ≥ PC1 < 0.306 indicates stage 4 chronic kidney disease (CKD). PC1 ≥ 0.3057 indicates CKD5 stage 5 chronic kidney disease.

6. A detection reagent, characterized in that, The set of markers obtained by screening using the marker set as described in claim 1 or the screening method as described in claim 2 is used as the detection target.

7. The product being tested, characterized in that, Includes the detection reagent as described in claim 6, as well as acceptable adjuvants, carriers, or devices.

8. The use of any one of the biomarker set as described in claim 1, the biomarker set obtained by the screening method as described in claim 2, the detection reagent as described in claim 6, or the detection product as described in claim 7 in the preparation of a kit for the prognosis and / or staging of chronic kidney disease and / or in the construction of a system for the prognosis and / or staging of chronic kidney disease.

9. A system for the prognosis and / or staging of chronic kidney disease, characterized in that, include: Data collection module, data analysis module, and results output module; The data collection module acquires the age of the subject from the biomarker set as described in claim 1, and the CD3 concentration in the subject's blood sample. + CD8 + .MMP low # Data on hemoglobin, serum creatinine, and serum potassium levels; The data analysis module: based on the subject's age and the CD3 levels in the subject's blood sample... + CD8 + .MMP low #, hemoglobin, serum creatinine and serum potassium levels, using the staging model in the application as described in any one of claims 3 to 5, calculate the immune nephropathy index and / or PC1; The result output module stages chronic kidney disease based on the magnitude of the immune nephropathy index or PC1 value.

10. The use of any one of the biomarker set as described in claim 1, the biomarker set obtained by the screening method as described in claim 2, the detection reagent as described in claim 6, the detection product as described in claim 7, or the system as described in claim 9 in an apparatus for preparing prognoses and / or staging of chronic kidney disease.

Images