Application of circulating protein FGF23 in predicting sudden cardiac death risk of chronic kidney disease patient and model construction

Abstract

The invention provides application of circulating protein FGF23 in predicting sudden cardiac death risk of chronic kidney disease patients and model construction, and belongs to the technical field of medicines. The biomarkers comprise NTproBNP (N-terminal B-type natriuretic peptide), DTNB (heteronectin beta), ANGPT2 (angiopoietin 2), FGF23 (fibroblast growth factor 23) and SEPTIN8 (cytokinin 8), wherein the circulating protein FGF23 has the best prediction effect after the risk ratio and AUC evaluation are combined. The FGF23 marker is provided for early prediction of sudden cardiac death after chronic kidney disease, and a technical means is provided for prognosis evaluation of chronic kidney disease treatment.

Description

Technical Field

[0001] The present invention belongs to the field of medical technology, and specifically relates to the application of circulating protein FGF23 in predicting the risk of sudden cardiac death in patients with chronic kidney disease and model construction. Background Art

[0002] Chronic kidney disease (CKD) refers to abnormal kidney structure or function caused by multiple factors that persists for more than 3 months. It has become a global health problem with high incidence, difficulty in detection, poor prognosis and expensive treatment. Epidemiology shows that the number of adult CKD patients in China ranks first in Asia. Early detection, targeted prevention, and minimization of controllable risk factors are currently the most effective measures to reduce the incidence of CKD. As one of the strongest risk factors for cardiovascular disease (CVD), CKD causes significantly more cardiovascular damage than CKD itself. 50% of CKD patients will have cardiovascular complications, which ultimately leads to a shortened life expectancy. CKD patients bear multiple burdens of cardiovascular metabolic and kidney disease risk factors and have a poor prognosis. In addition, in addition to cardiometabolic disorders, the potential risks of CKD patients receiving treatment should not be underestimated. Although renal treatment can improve renal function in the short term, it is still uncertain whether renal treatment can reduce the long-term CVD risk of CKD patients. Some treatments are even believed to aggravate metabolic disorders in CKD patients, leading to increased CVD and mortality.

[0003] Sudden cardiac death (SCD) is the leading cause of death from CVD, and the mortality rate due to SCD is reported to be >60% in patients with advanced CKD. However, limited attention has been paid to the association between CKD and SCD risk. Given the rapid onset and high lethality of SCD, it is important to identify the risk of SCD in patients with CKD. Therefore, it is crucial to construct a model to predict the occurrence of SCD after CKD.

[0004] FGF23 (Fibroblast Growth Factor 23) is a functional protein molecule containing 251 amino acid residues. It is proteolytically cleaved into an N-terminal and a C-terminal fragment between arginine at position 179 and serine at position 180 of the protein molecule. FGF23 is a 32 kDa secreted circulating protein, mainly produced by osteocytes in bone, and acts in distant target organs in an endocrine manner. FGF23 usually requires fibroblast growth factor receptor (FGFR) and α-Klotho to initiate its signal transduction. Elevated circulating FGF23 levels and reduced Klotho protein levels can be detected in the early stage of CKD, and a large number of studies have shown that FGF23 and Klotho protein are involved in the pathological process of cardiovascular complications in CKD. However, there is currently no study exploring the role of circulating protein FGF23 in predicting the risk of SCD in CKD patients and constructing a corresponding prediction model. Summary of the Invention

[0005] In view of this, the present invention addresses the problems existing in the prior art and provides the application of circulating protein FGF23 in predicting the risk of sudden cardiac death in patients with chronic kidney disease and the construction of a model.

[0006] To achieve the above object, the technical solutions adopted by the present invention are as follows:

[0007] On the one hand, the present invention provides a biomarker for constructing a prediction model for sudden cardiac death in patients with chronic kidney disease. The biomarker is circulating protein FGF23, which has the protein sequence shown in SEQ ID NO.1.

[0008] SEQ ID NO.1:

[0009] MLGARLRLWVCALCSVCSMSVLRAYPNASPLLGSSWGGLIHLYTATARNSYHLQIHKNGHVDGAPHQTIYSALMIRSEDAGFVVITGVMSRRYLCMDFRGNIFGSHYFDPENCRFQHQTLENGYDVYHSPQYHFLVSLGRAKRAFLPGMNPPPYSQFLSRRNEIPLIHFNTPIPRRHTRSAEDDSERDPLNVLKPRARMTPAPASCSQELPSAEDNSPMASDPLGVVRGGRVNTHAGGTGPEGCRPFAKF.

[0010] Specifically, the risk model for sudden cardiac death after chronic kidney disease is constructed with circulating protein FGF23 as the dependent variable.

[0011] Specifically, the risk determination criterion with the circulating protein FGF23 as the dependent variable is as follows: When the normalized level of the circulating protein FGF23 > 0.2, it is determined as sudden cardiac death after chronic kidney disease.

[0012] Preferably, the risk determination criterion with the circulating protein FGF23 as the dependent variable is as follows: When the normalized level of the circulating protein FGF23 > 0.209, it is determined as sudden cardiac death after chronic kidney disease.

[0013] Among them, after the data is first processed by quality control, the differential protein level is converted into a normalized protein expression (NPX) value, and the values with a protein level greater than 5 times the median absolute deviation are directly deleted.

[0014] On the other hand, the present invention provides a screening method for a sudden cardiac death prediction model for chronic kidney disease patients based on the circulating protein FGF23, including the following steps:

[0015] S1. Selection of the modeling population: Select the UK Biobank and the Changsha cohort in China that meet the inclusion criteria of the study group;

[0016] S2. Information collection: Collect the plasma sample tests of the modeling population;

[0017] S3. Establishment of the prediction model: Use the selected UK Biobank and Changsha cohort in China of the modeling population as the training set and the validation set respectively to construct a sudden cardiac death prediction model after chronic kidney disease;

[0018] S4. Evaluation of the model prediction ability: Evaluate the prediction ability of the sudden cardiac death prediction model constructed in step S3.

[0019] Specifically, the information collection in step S2 also includes covariates, and the covariates are demographic characteristics and routine clinical medicine characteristics.

[0020] More specifically, the routine clinical medicine characteristics include any one or more of body mass index, systolic blood pressure, diastolic blood pressure, total cholesterol, serum creatinine, glomerular filtration rate, urinary albumin creatinine ratio, hypertension at baseline, diabetes, hyperlipidemia, and cardiovascular disease history.

[0021] Specifically, based on the risk determination with the circulating protein FGF23 as the dependent variable, the risk is higher when the age ≥ 60 years old; the risk is higher for men.

[0022] Specifically, the risk is higher for patients with hypertension; the risk is higher for patients with diabetes; the risk is higher for patients with a history of cardiovascular disease.

[0023] Specifically, the screening of plasma differential proteins is also included in step S4, and only the differential proteins with P FDR <0.05 meet the requirements.

[0024] On the other hand, the present invention provides an application of the biomarker FGF23 in preparing a prediction model for sudden cardiac death in patients with chronic kidney disease.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] The application of the circulating protein FGF23 disclosed in the present invention in predicting the risk of sudden cardiac death in patients with chronic kidney disease and the construction of the model. In predicting the risk of SCD occurring in five years, the predictive AUC of FGF23 is greater than 0.7. The other proteins screened may also be involved in the diagnostic markers of SCD, which provides a new technical means for the early diagnosis and prognosis evaluation of sudden cardiac death after chronic kidney disease. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0028] Figure 1 In A is the screening result of differential proteins related to CKD and SCD, and in B is the cross-validation result of the Lasso-Cox model. In B, the "upper abscissa" represents the number of variables corresponding to different Log(λ).

[0029] Figure 2 It is the prediction accuracy result of the key protein SCD risk. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0030] The various exemplary embodiments of the present invention will now be described in detail. This detailed description should not be considered as a limitation of the present invention, but rather as a more detailed description of certain aspects, characteristics, and implementation embodiments of the present invention. It should be understood that the terms described in the present invention are only for describing specific implementation embodiments and are not used to limit the present invention. Additionally, for the numerical ranges in the present invention, it should be understood that each intermediate value between the upper and lower limits of the range is also specifically disclosed. Each intermediate value within any stated value or stated range, as well as each smaller range between any other stated value or intermediate value within the stated range, is also included in the present invention. The upper and lower limits of these smaller ranges may be independently included or excluded from the range.

[0031] Example 1 Construction of a risk prediction model for sudden cardiac death after chronic kidney disease

[0032] (I) Construction of a risk prediction model for sudden cardiac death after chronic kidney disease, specifically the following steps:

[0033] S1. Modeling population selection: The UK Biobank (discovery cohort) and the Changsha cohort (validation cohort) in China that met the inclusion criteria of the study group were selected, and individuals with missing relevant covariate information and patients diagnosed with transient SCD before baseline were excluded. The follow-up ended on December 30, 2021. The Changsha cohort included was people aged 18 years or older who underwent physical examinations at the Health Management Center of the Third Xiangya Hospital from January 1, 2009 to December 31, 2020. The death information tracking time was from January 1, 2009 to February 24, 2021;

[0034] S2. Information collection: The following information of the modeling population was collected: demographic characteristics, routine clinical medical characteristics, and plasma sample testing; demographic characteristics included age, gender, and race; routine clinical medical characteristics included body mass index (BMI), systolic blood pressure, diastolic blood pressure, total cholesterol, serum creatinine, glomerular filtration rate, urine albumin-to-creatinine ratio, hypertension, diabetes, hyperlipidemia, and cardiovascular disease history at baseline; and also included socioeconomic status (average household income) and lifestyle (smoking, drinking, cholesterol-lowering drug treatment, blood pressure, and diabetes) information collected through a self-administered questionnaire.

[0035] Assessment of CKD: Baseline eGFR was calculated using serum creatinine in the discovery cohort according to the 2021 CKD-EPI race-free equation. Individuals were considered to have CKD if they met any of the following criteria: ① CKD diagnosed according to ICD-10 codes, including N18.0, N18.1, N18.2, N18.3, N18.4, N18.5, N18.9, Z99.2, Z49.0, and Z49.1; ② estimated glomerular filtration rate (eGFR) < 60 mL / min / 1.73 m 2 ; ③ Baseline urinary albumin / creatinine ratio (UACR) ≥ 30 mg / g. Due to data limitations, in the validation cohort, we only considered patients with baseline eGFR < 60 mL / min / 1.73 m 2 Diagnosis of CKD.

[0036] Assessment of SCD: In the discovery cohort, the outcome of this study was SCD, including survival of patients diagnosed with SCD and death due to SCD, which is ICD-10 codes I46 and I49. In the validation cohort, the outcome was death due to SCD, including the primary or secondary cause of death being listed as sudden cardiac death or cardiac arrest.

[0037] S3. Risk prediction model establishment: The selected modeling populations, the UK Biobank and the Changsha cohort in China, were used as the training set and the validation set respectively to construct a prediction model for sudden cardiac death after chronic kidney disease;

[0038] To evaluate the outcome differences between participants with and without CKD, a Cox proportional hazards model was used to analyze the relationship between CKD and the risk of SCD. We constructed two models: a model adjusted for age, sex, and race (Model 1), and another model adjusted for obesity, baseline eGFR, lifestyle factors, socioeconomic status, history of hypertension, diabetes, hyperlipidemia, and history of cholesterol, blood pressure, and diabetes medications (Model 2). In the validation cohort, Model 1 was adjusted for age and sex; Model 2 was adjusted for obesity, baseline eGFR, lifestyle factors, socioeconomic status, history of hypertension, diabetes, hyperlipidemia, and cholesterol, blood pressure, and diabetes medications.

[0039] (2) Validation of the risk prediction model for sudden cardiac death after chronic kidney disease

[0040] After information collection, it was found that the median age of the study subjects in the discovery cohort was 57 years (IQR, 49 - 63 years), and 48.91% were male. 4.83% had a history of diabetes, 7.12% had a history of hypertension, and 10.92% had a history of CVD. The median age of the study subjects in the validation cohort was 40 years (IQR, 31 - 50 years), and there were 73,029 males (65.72%). The histories of diabetes, hypertension, and CVD accounted for 8.66%, 16.90%, and 15.82% respectively. Compared with the discovery cohort, the population in the validation cohort was younger, had a higher proportion of males, and had higher proportions of histories of diabetes and hypertension.

[0041] During a median follow-up of 13.63 years (IQR, 12.88 - 14.35 years), a total of 6,815 study subjects had SCD. In the validation cohort, we observed 154 SCD cases (median follow-up time was 8 years). In the discovery cohort, compared with patients without CKD, the HR for SCD in CKD patients in Model 1 was 1.81 (95% CI, 1.68 - 1.96, P < 0.001). The HR value of Model 2 was 1.33 (95% CI, 1.22 - 1.45, P < 0.001). In the validation cohort, the adjusted HRs for SCD in CKD patients in Model 1 and Model 2 were 2.08 (95% CI, 1.37 - 3.17, P < 0.001) and 1.87 (95% CI, 1.05 - 3.33, P = 0.034) respectively. These results indicate that in both cohorts, the risk of SCD in CKD patients was relatively high, suggesting that the data from both cohorts can be used to screen for risk prediction markers.

[0042] Example 2 Accuracy of circulating protein FGF23 in predicting SCD in CKD patients

[0043] (I) Screening of differentially expressed proteins in plasma samples of CDK and non-CDK patients

[0044] In Example 1, all the final cohort populations with protein data were used as the training set. In the training set, it was divided into the CKD group and the non-CKD group. First, the differential proteins between CKD and non-CKD patients were analyzed using the t-test. Only proteins with P FDR <0.05 met the requirements. The results are as shown in Figure 1 A. A total of 1250 differentially expressed proteins were identified. In the differential protein analysis, if the protein was missing in >50% of the population, the average value of the protein was used for analysis. If the protein was missing in >80% of the population, it was excluded from the analysis; further multiple linear regression analysis was used. After Bonferroni correction, 1335 proteins were found to be significantly associated with CKD. Cox proportional hazards model analysis further identified 92 proteins related to the risk of SCD. Finally, the overlapping proteins were analyzed by the Lasso-Cox model and cross-validated. The results are as shown in Figure 1 B. The cross-validation error was small and the data was reliable. Five candidate proteins that met the criteria of minimizing the average cross-validation error were screened out, including NTproBNP, DTNB, ANGPT2, FGF23, and SEPTIN8.

[0045] (II) Accuracy of differential proteins in predicting SCD

[0046] In Example 1, all CKD patients with protein data were selected as the validation set (N = 2280). Since in the protein analysis of the validation set, the data of DTNB and SEPTIN8 were less than 60% of the total sample size of CKD patients, these two proteins were removed from the Cox proportional hazards model validation and ROC curve analysis. Figure 2 The hazard ratio (HR) and P-value of the final model of the 3 key proteins related to SCD risk using the Cox proportional hazards model in the validation set were shown, as well as the predicted AUC of the key proteins in the model for predicting the 5-year SCD incidence. Considering that FGF23 had the largest HR risk ratio in the validation set and an AUC > 0.7 for 5 years, it was considered that FGF23 had good predictive ability for the SCD risk in CKD patients. When the level of circulating protein FGF23 after normalization > 0.209, it was determined as sudden cardiac death after chronic kidney disease.

[0047] Finally, it should be noted that the above content is only used to illustrate the technical solution of the present invention, rather than limiting the protection scope of the present invention. Any simple modification or equivalent replacement made by those of ordinary skill in the art to the technical solution of the present invention shall not depart from the essence and scope of the technical solution of the present invention.

Claims

1. A biomarker for constructing a prediction model for sudden cardiac death in patients with chronic kidney disease, characterized in that: The biomarker is a circulating protein FGF23, having a protein sequence as shown in SEQ ID NO.

1.

2. The biomarker according to claim 1, characterized in that The risk model for sudden cardiac death after chronic kidney disease was constructed using the circulating protein FGF23 as the dependent variable.

3. The biomarker according to claim 2, characterized in that The risk determination criterion for using the circulating protein FGF23 as a dependent variable is: when the normalized level of circulating protein FGF23 is > 0.2, it is determined to be sudden cardiac death after chronic kidney disease.

4. The biomarker screening method according to claim 1, characterized in that: The screening method comprises the following steps: S1. Modeling population selection: The UK Biobank and Changsha cohort in China that met the inclusion criteria of the study group were selected; S2, information collection: collect plasma samples of the modeling population for testing; S3. Prediction model establishment: The selected modeling populations, the UK Biobank and the Changsha cohort in China, were used as training sets and validation sets, respectively, to construct a prediction model for sudden cardiac death after chronic kidney disease; S4. Model prediction ability evaluation: Evaluate the prediction ability of the sudden cardiac death prediction model after chronic kidney disease constructed in step S3.

5. The screening method according to claim 4, characterized in that The information collection in step S2 also includes covariates, which are demographic characteristics and routine clinical medical characteristics.

6. The screening method according to claim 5, characterized in that The demographic characteristics include any one or more of age, gender, and race.

7. The screening method according to claim 5, characterized in that The conventional clinical medical characteristics include any one or more of body mass index, systolic blood pressure, diastolic blood pressure, total cholesterol, serum creatinine, glomerular filtration rate, urine albumin-to-creatinine ratio, hypertension at baseline, diabetes, hyperlipidemia, and cardiovascular disease history.

8. The screening method according to claim 4, characterized in that Based on the risk determination using the biomarkers as dependent variables, those aged > 60 years have a higher risk; and those who are male have a higher risk.

9. The screening method according to claim 7, characterized in that People with high blood pressure are at higher risk; people with diabetes are at higher risk; and people with a history of cardiovascular disease are at higher risk.

10. Use of the biomarker according to any one of claims 1 to 3 or the biomarker obtained by the screening method according to claim 4 in preparing a prediction model for sudden cardiac death in patients with chronic kidney disease.

Images