Use of a biomarker combination in constructing a predictive model for intracranial aneurysm rupture

Abstract

The application provides an application of a peripheral blood biomarker combination in construction of a prediction model for intracranial aneurysm rupture, solves the problems of the existing intracranial unruptured aneurysm prediction method and the lack of related early warning biomarkers in the prediction model, and the technical scheme is as follows: an application of a biomarker combination in construction of a prediction model for intracranial aneurysm rupture, the biomarker combination is factor: GPC1, ECM2 and PLEC, a prediction model for predicting intracranial aneurysm rupture is constructed based on the biomarker, and the application and application method in detection reagents and kits have the advantages and positive effects that the rupture of an aneurysm is predicted early, the pain of a patient is reduced, and a reference basis is provided for individualized diagnosis and treatment; the prediction model constructed by the biomarkers has the characteristics of high sensitivity and good specificity, and is significantly better than the PHASES scoring method and the unruptured intracranial aneurysm treatment scoring method.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of biomedical technology, in particular to application of a biomarker combination in constructing a prediction model for rupture of intracranial aneurysm, early warning marker for predicting rupture of intracranial aneurysm and new use of the prediction model. BACKGROUND

[0002] Intracranial aneurysm is the result of multiple factors leading to arterial wall degeneration and abnormal expansion. The incidence of intracranial aneurysm in the general population is about 3.6-7%, and the rupture rate is about 1-2%. Once ruptured, intracranial aneurysm can cause subarachnoid hemorrhage, and the mortality rate of the first hemorrhage is more than 25%, and the disability rate of the survivors is as high as about 50%. About 85% to 90% of intracranial aneurysm patients are asymptomatic. With the development of imaging technology, the detection rate of small intracranial aneurysms (diameter ≤7mm) has significantly improved. In the past, it was believed that small aneurysms had low rupture risk and should be treated conservatively. However, subsequent studies found that about 71.8% of ruptured aneurysms were small aneurysms. Currently, 66% of intracranial aneurysm patients in adults are small aneurysms, and there is no consensus on the timing of surgery for these asymptomatic small aneurysms. Current aneurysm surgical methods, whether endovascular embolization or micro-clip occlusion, can cause postoperative complications, so the risk and benefit of surgery are uncertain. A considerable number of patients adopt a conservative approach, but there is also a certain risk of rupture. Therefore, early judgment of the stability and rupture risk of asymptomatic small aneurysms and selection of the best timing of surgery are very important. Individualized assessment of aneurysm rupture risk is crucial for weighing the risks of treatment and observation.

[0003] Currently, internationally established clinical scoring models for predicting aneurysm rupture and aneurysm growth risk include (1) PHASES (P race, H hypertension history, A age, S aneurysm size, E aneurysmal subarachnoid hemorrhage history, S aneurysm location) scoring method; (2) unruptured intracranial aneurysm treatment scoring method. These two scoring methods have different focuses, both of which score according to age, race, aneurysm location, size and shape, history of hypertension, history of subarachnoid hemorrhage, etc., each with its own drawbacks. Among them, the accuracy of PHASES scoring method needs to be improved, and aneurysms less than 7mm are defined as low risk, but actual clinical findings show that small aneurysms have a high risk of rupture, which may be biased for the prediction of small aneurysms. The unruptured intracranial aneurysm treatment scoring method includes 29 variables, which is too complex to be widely used.

[0004] In the process of intracranial aneurysm formation, the blood vessel wall is locally deteriorated, the aneurysm wall is pathologically changed, the smooth muscle cells are transformed into pro-inflammatory phenotype, the mucous material is deposited in the vascular intima and mesoderm, and the elastic plate is broken to cause the weakness of the arterial wall elasticity. The loss and foaming of the arterial wall cells eventually lead to the rupture of the aneurysm. The series of pathological changes of the aneurysm wall lead to the release of some marker characteristic molecules into the peripheral blood. If the early warning biomarker related to aneurysm rupture can be found in the peripheral blood, a reliable reference basis for the operation time of unruptured aneurysm will be provided, which will be a major progress in the field. SUMMARY

[0005] The purpose of the present application is to provide an application of a peripheral blood biomarker combination in constructing a prediction model of intracranial aneurysm rupture, to solve the problems of the existing unruptured intracranial aneurysm prediction method and the lack of related early warning biomarker application in the prediction model, to combine three factors as biomarker combination in the prediction model, and to provide a new use of the biomarker combination as a detection reagent and a kit.

[0006] The technical solution of the present application is: an application of a biomarker combination in constructing a prediction model of intracranial aneurysm rupture, and the technical points are: the biomarker combination is composed of factors: GPC1, ECM2 and PLEC, and a prediction model of intracranial aneurysm rupture is constructed based on the biomarker.

[0007] The expression formula of the prediction model is:

[0008] ln p / (1-p) = 1.7640*C GPC1 +0.0134*C ECM2 +0.1775*C PLEC +(-41.7000)

[0009] Wherein C GPC1 , C ECM2 and C PLEC are respectively GPC1 concentration, ECM2 concentration and PLEC concentration in the peripheral blood plasma of the sample, and the values are the specific measured values, and the aneurysm rupture risk probability p of the sample is calculated.

[0010] The present application also provides an application of the biomarker combination in constructing a prediction model of intracranial aneurysm rupture in a detection reagent, and the technical points are: the detection reagent is composed of reagents for detecting the plasma expression amount of biomarker combination factors GPC1, ECM2 and PLEC.

[0011] The application also provides an application method of the biomarker combination in constructing the prediction model of intracranial aneurysm rupture, and a technical point is that a nomogram is constructed according to the biomarker combination in constructing the prediction model of intracranial aneurysm rupture, a rupture risk probability is evaluated by the nomogram scoring, a plasma GPC1 concentration value range, a plasma PLEC concentration value range and a plasma ECM2 concentration value range number axis are respectively set, a risk score number axis, a risk total score number axis and a rupture risk probability number axis are respectively set, risk score values of the respective plasma concentrations of GPC1, PLEC and ECM2 are respectively taken from the risk score number axis, a total of the respective risk score values is calculated, then the total of the respective risk score values is found on the risk total score number axis, and finally the corresponding risk probability is found on the rupture risk probability number axis.

[0012] The application also provides an application method of the biomarker combination in constructing the prediction model of intracranial aneurysm rupture, and a technical point is that a nomogram is constructed according to the biomarker combination in constructing the prediction model of intracranial aneurysm rupture, a rupture risk probability is evaluated by the nomogram scoring, a plasma GPC1 concentration value range, a plasma PLEC concentration value range and a plasma ECM2 concentration value range number axis are respectively set, a risk score number axis, a risk total score number axis and a rupture risk probability number axis are respectively set, risk score values of the respective plasma concentrations of GPC1, PLEC and ECM2 are respectively taken from the risk score number axis, a total of the respective risk score values is calculated, then the total of the respective risk score values is found on the risk total score number axis, and finally the corresponding risk probability is found on the rupture risk probability number axis.

[0013] The application has the advantages and positive effects that:

[0014] (1) At present, the judgment of asymptomatic small aneurysm rupture in clinical practice is still in a state of lacking precise early warning biomarkers, and the application combines various frontier scientific and technological means such as isotope labeling relative and absolute quantification, high-performance liquid chromatography fractionation technology and liquid chromatography-mass spectrometry technology, performs proteomic analysis on plasma samples, finds biomarkers suitable for early warning of intracranial aneurysm rupture through high-throughput screening of differential proteins.

[0015] (2) The application uses peripheral blood plasma samples, is based on non-invasive diagnosis and treatment technology, early predicts the rupture of aneurysm, reduces the pain of patients, and provides a reference basis for individualized diagnosis and treatment.

[0016] (3) The application compares the protein expression differences in the plasma of unruptured intracranial aneurysm patients and ruptured aneurysm patients, uses traumatic subarachnoid hemorrhage patients and healthy people as a control group, excludes stress response factors in blood caused by aneurysm rupture and hemorrhage stimulation of meninges, thereby screening biomarkers related to aneurysm rupture, and successfully obtaining a prediction model for predicting intracranial aneurysm rupture.

[0017] (4) The prediction model constructed by the biomarkers disclosed in the application has high sensitivity and good specificity, the area under the ROC curve is 0.966 (95% confidence interval: 0.927-1.000), the sensitivity for distinguishing the ruptured and unruptured intracranial aneurysm population is 82.8%, and the specificity is 100%. The positive predictive rate of the prediction model for intracranial aneurysm rupture is 95.0%, which is significantly better than the PHASES scoring method and the unruptured intracranial aneurysm treatment scoring method (the positive predictive rates are 66.7% and 62.5%, respectively). BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 The experimental flowchart for the application of the application;

[0019] Figure 2 The screening diagram of the differentially expressed proteins of each group of the application;

[0020] Figure 3 The plasma expression level diagram of GPC1, ECM2 and PLEC of each group of the application;

[0021] Figure 4 The nomogram of the prediction model of the application;

[0022] Figure 5 The ROC curve diagram of the training set of the application;

[0023] Figure 6 The ROC curve diagram of the verification set of the application;

[0024] Figure 7 The ROC curve diagram of the plasma GPC1, ECM2 and PLEC protein concentration predicting aneurysm rupture of the application;

[0025] Figure 8 The ROC curve diagram of the plasma GPC1xPLEC, ECM2xPLEC and GPC1xECM2 combined prediction of aneurysm rupture of the application. EMBODIMENT

[0026] The application provides a new use of a biomarker combination, that is, as an application in constructing a prediction model for intracranial aneurysm rupture. The application will be further described in detail below in combination with specific examples. The materials, reagents, drugs, software systems and the like used in the examples can be obtained from commercial channels unless otherwise specified.

[0027] The biomarker combination used is factor: GPC1, ECM2 and PLEC, wherein GPC1 is Glypican-1, ECM2 is Extracellular matrix protein 2, and PLEC is Plectin. A prediction model for predicting the rupture of intracranial aneurysm is constructed based on the biomarkers.

[0028] The screening of the biomarkers of the application comprises the following steps:

[0029] (I) Experimental method

[0030] 1. Experimental grouping

[0031] Four groups were set up in the present study, namely, the aneurysm rupture group (RIA group), the aneurysm unrupture group (UIA group), the traumatic subarachnoid hemorrhage control group (Traumatic SAH Control group, tSAH), and the healthy control group (Healthy Control, HC group).

[0032] 2. Inclusion criteria and process

[0033] As shown in the flowchart of the sample inclusion process of the present application. Figure 1

[0034] (1) Inclusion criteria of aneurysm patients (RIA & UIA): Patients confirmed to have aneurysm by digital subtraction angiography or CT angiography.

[0035] (2) Inclusion criteria of tSAH group: Patients with non-vascular subarachnoid hemorrhage caused by trauma, confirmed by CT imaging to have no aneurysm.

[0036] (3) Inclusion criteria of HC group: Good physical health, no abnormal values in blood routine, liver function, kidney function, etc.

[0037] (4) Exclusion criteria: Cavernous angioma; subarachnoid hemorrhage caused by non-intracranial aneurysm factors (vascular malformations, etc.); dissecting aneurysm; malignant tumors, autoimmune diseases and other systemic diseases; patients who have received surgical, interventional and other surgical treatments, etc.

[0038] 3. Sample collection

[0039] The enrolled population was sampled with 5 milliliters of elbow venous blood within 2 hours of physical examination or hospitalization. The sample was collected in an anticoagulant blood collection tube, centrifuged at 3500 revolutions per minute for 10 minutes at 4 degrees Celsius, and the plasma was aliquoted and stored at -80 degrees Celsius at 0.2 milliliters per tube until analysis.

[0040] ​4. Isobaric and absolute quantification and liquid chromatography-tandem mass spectrometry analysis

[0041] 0.2 mL of each patient's plasma sample was taken for thorough mixing, and 10 patient's plasma mixed sample was taken for each group. High abundant proteins were removed from the plasma samples using ProteoMiner™ Protein Enrichment Kit (Bio-Rad, USA), and the treated plasma was determined for protein concentration using BCA Protein Assay Kit. The peptides obtained after trypsin digestion of the proteins were desalted using C18 extraction cartridges, and the peptides were vacuum centrifuged and dried. The dried peptides were reconstituted in 40 μL of 0.1% formic acid. 100 μg of the peptide mixture was taken for each sample, and the peptides were labeled using the Isobaric and Absolute Quantification Kit, and the labeled peptides were separated by SCX chromatography using an AKTA Purification System (GE Healthcare). The dried peptide mixture was reconstituted and acidified with buffer A (10 mM potassium dihydrogen phosphate in acetonitrile water solution, acetonitrile 25%, pH = 3.0) and fractionated on a 4.6 x 100 mm column (5 μm, 200 A, Poly LC Inc.) at a flow rate of 1 mL / min. The absorbance at 214 nm was monitored during the elution, and the eluted fractions were collected every 1 min. The fractions were desalted using C18 Cartridge columns and vacuum centrifuged and lyophilized. Each sample was separated using a nanoflow high performance liquid system, Easy nLC (Thermo Scientific). The column was equilibrated with 95% of buffer A (0.1% formic acid). The sample was first loaded on a pre-column (Thermo Scientific Acclaim PepMap 100, 100 μm x 2 cm, nanoViper C18), and then separated by linear elution using a C18 reverse phase analytical column (Thermo Scientific Easy Column, 75 μm x 10 cm, 3 μm). The eluent was buffer B (0.1% formic acid in acetonitrile water solution, acetonitrile 84%) at a flow rate of 300 nL / min. The separated peptides were analyzed using an Orbitrap mass spectrometer (Thermo Scientific).

[0042] 5. Protein identification and quantification

[0043] The proteins were identified and quantified using Proteome Discoverer 1.4 software embedded with MASCOT engine (version 2.2). The ratio of the protein expression level in the RIA, UIA and tSAH groups to that in the HC group was taken as the differential fold of each protein. When the differential fold changed by >1.2 fold or <0.83 fold, and P < 0.05, the protein was considered to be a differentially expressed protein.

[0044] 6. Enzyme-linked immunosorbent assay

[0045] Plasma protein concentration was determined using enzyme-linked immunosorbent assay (ELISA). The test reagents were purchased from Shanghai Enzyme-Linked Biotechnology Co., Ltd., and the test process was strictly carried out in accordance with the instructions.

[0046] 7. Statistical methods

[0047] Statistical analysis and graphing were performed using SPSS 23.0 and the rms and pROC packages in R version 4.2.2. Quantitative data were tested for normality using the Kolmokolov-Smilov test. Normally distributed quantitative data were expressed as mean ± standard deviation, and comparisons between groups were performed using Student's t-test or one-way ANOVA. Non-normally distributed quantitative data were expressed as median (interquartile range), and comparisons between groups were performed using the Mann-Whitney rank-sum test or the Kruskal-Wallis test. Categorical data were expressed as percentages, and comparisons between groups were performed using the chi-square test or Fisher's exact test. Receiver operating characteristic (ROC) curves were used to calculate the predictive power of variables, estimating the area under the curve (AUC) and the corresponding 95% confidence intervals. The statistical significance level was set at 0.05.

[0048] 8. Construction of a nomogram model for predicting intracranial aneurysm rupture

[0049] The odds ratios of the 95% confidence intervals were calculated using univariate and multivariate logistic regression to exclude confounding variables. Aneurysm rupture-related variables with p < 0.05 in the univariate logistic regression analysis were included in the discussion of the multivariate regression analysis. If these factors still maintained a significant correlation with p < 0.05, they were included in the predictive model, and then nomograms were constructed using these variables.

[0050] (II) Experimental Results

[0051] 1. Analysis of detection results using isotope-labeled relative and absolute quantitative methods combined with liquid chromatography-mass spectrometry.

[0052] A total of 859,140 secondary spectra were obtained by isotope labeling relative and absolute quantification combined with liquid chromatography-mass spectrometry, of which 43,468 were database-matched; a total of 6,624 peptides were identified, of which 5,672 were unique peptides; and 812 proteins were identified, of which 798 were quantifiable.

[0053] 2. Screening of differentially expressed proteins associated with aneurysm formation or rupture

[0054] like Figure 2As shown, proteins exhibiting significant changes only in the RIA or UIA groups were selected as differentially expressed proteins. If a protein showed a significant change in the tSAH control group along the same trend as in RIA or UIA, it was considered not a specific differentially expressed protein. According to the Venn diagram results, 57 differentially expressed proteins showed significant changes only in UIA patients, with 27 proteins showing significantly increased expression and 30 proteins showing significantly decreased expression. Figure 2 As shown in the abe region. 43 differentially expressed proteins showed significant changes only in RIA patients, with 32 proteins showing significantly increased expression and 11 proteins showing significantly decreased expression, such as... Figure 2 As shown in the cdef region.

[0055] Analysis of the results of relative and absolute quantitative analysis using isotope labeling combined with liquid chromatography-mass spectrometry showed that GPC1, ECM2, and PLEC were differentially expressed proteins with significantly upregulated relative expression levels only in the plasma of patients in the RIA group.

[0056] 3. Expression of GPC1, ECM2, and PLEC in patient plasma

[0057] The concentrations of plasma biomarkers in each group were detected using an enzyme-linked immunosorbent assay (ELISA), and the results are as follows: Figure 3 As shown.

[0058] The mean plasma GPC1 concentration was (5.44±2.18) ng / ml in the RIA group, (3.36±1.47) ng / ml in the UIA group, (4.05±1.17) ng / ml in the traumatic subarachnoid hemorrhage control group, and (3.03±0.88) ng / ml in the healthy control group. One-way ANOVA showed that the plasma GPC1 concentration in the rupture group was significantly different from the other three groups (P<0.001), with a significantly higher concentration. There were no significant differences between the unruptured group and the tSAH control group and the HC group (P=0.303 and P=0.697, respectively).

[0059] The mean plasma ECM2 concentration was (1721.38±528.07) ng / ml in the RIA group, (1183.60±420.47) ng / ml in the UIA group, (1380.83±449.82) ng / ml in the traumatic subarachnoid hemorrhage control group, and (953.84±482.92) ng / ml in the healthy control group. One-way ANOVA showed that the plasma ECM2 concentration in the rupture group was significantly different from the other three groups (P<0.001), with a significantly higher ECM2 concentration in the rupture group. There were no significant differences between the unruptured group and the tSAH control group and the HC group (P=0.280 and P=0.321, respectively).

[0060] The mean plasma PLEC concentration was (87.67±15.87) ng / ml in the RIA group, (64.81±20.11) ng / ml in the UIA group, (71.32±18.21) ng / ml in the tSAH group, and (57.82±21.32) ng / ml in the HC group. Univariate radiotherapy analysis showed that the plasma PLEC concentration in the RIA group was significantly different from that in the other three groups (P<0.001), indicating a significantly higher PLEC concentration in the RIA group. There were no significant differences between the UIA group and the tSAH control group and the HC group (P=0.355 and P=0.433, respectively).

[0061] 4. Peripheral blood GPC1 plasma levels are an independent risk factor for intracranial aneurysm rupture.

[0062] The following is a list of risk factors for intracranial aneurysm rupture predicted by univariate and multivariate logistic regression models:

[0063]

[0064] Univariate logistic regression analysis showed that the odds ratio (ORR) between plasma GPC1 levels and intracranial aneurysm rupture was 2.438 (95% confidence interval: 1.545–2.324, P < 0.001), the ORR between plasma ECM2 levels and intracranial aneurysm rupture was 1.010 (95% confidence interval: 1.004–1.016, P < 0.001), and the ORR between plasma PLEC levels and intracranial aneurysm rupture was 1.157 (95% confidence interval: 1.057–1.267, P = 0.002). Incorporating these indicators into a multivariate logistic regression model, the results showed that GPC1, ECM2, and PLEC were all independent risk factors for aneurysm rupture.

[0065] 5. Construction, use and validation of nomogram model

[0066] The 244 patients with intracranial aneurysms enrolled in the study were randomly divided into the experimental group (70%, n=171) and the validation group (30%, n=73). By comparing the basic indicators of the patients in the experimental group and the validation group, there was no significant statistical difference between the experimental group and the validation group, indicating that the grouping was reliable. According to the logistic regression results, we used the independent risk factors (including GPC1 plasma concentration, ECM2 plasma concentration and PLEC plasma concentration) in the multivariate logistic regression analysis to jointly establish a clinical prediction model for intracranial aneurysm rupture, and visualize it through a nomogram. The formula for calculating the risk probability p obtained by the prediction model is:

[0067] ln p / (1-p) = 1.7640*C GPC1 + 0.0134*C ECM2 +0.1775*C PLEC +(-41.7000)

[0068] Where C GPC1 , C ECM2 and C PLEC are the plasma protein expression concentrations of the patient samples, and are assigned to specific measured values.

[0069] As shown in Figure 4 , the above model is visualized in the form of a nomogram: a method for applying a biomarker combination to construct a prediction model for intracranial aneurysm rupture, a nomogram is established according to the biomarker combination to construct a prediction model for intracranial aneurysm rupture, and the rupture risk probability is evaluated by the nomogram score; the plasma GPC1 concentration value range, the plasma PLEC concentration value range and the plasma ECM2 concentration value range are set respectively, the risk score axis, the risk total score axis and the rupture risk probability axis; the risk score axis corresponding to the plasma concentration value of GPC1, PLEC and ECM2 is respectively taken as the risk score value, the sum of each risk score value is calculated, then the corresponding risk score value sum is found on the risk total score axis, and finally the corresponding risk probability is found on the rupture risk probability axis.

[0070] A plasma GPC1 level of 0 ng / ml corresponds to a risk score of 0; a plasma GPC1 level of 10 ng / ml corresponds to a risk score of 66. A plasma PLEC level of 10 ng / ml corresponds to a risk score of 0; a plasma PLEC level of 110 ng / ml corresponds to a risk score of 66. A plasma ECM2 level of 0 ng / ml corresponds to a risk score of 0; a plasma ECM2 level of 2000 ng / ml corresponds to a risk score of 100. When using this method, based on the specific plasma concentration, move upwards to the first row, the "Risk Score" axis, to find the corresponding risk score. Calculate the sum of the scores for each factor. Find the total risk score in the "Total Risk Score" row. A total risk score of 132 corresponds to a risk probability of 0.01; a total risk score of 149 corresponds to a risk probability of 0.5; and a total risk score of 166 corresponds to a risk probability of 0.99. Move downwards to the "Rupture Risk Probability" row to find the corresponding risk probability of rupture for patients with intracranial aneurysms.

[0071] The ROC curve analysis model was used to distinguish between RIA and UIA. Ten-fold cross-validation was performed within the experimental group. The mean area under the curve (AUC) reached 0.953 (95% confidence interval: 0.904–1.000), with a sensitivity of 82.7% and a specificity of 98.1%. Figure 5 In the validation group, the area under the curve reached 0.966 (95% confidence interval: 0.927–1.000), with a sensitivity of 82.8% and a specificity of 100%. Figure 6 The prompt has good distinguishability.

[0072] The sensitivity and specificity of GPC1, ECM2, and PLEC as biomarkers were evaluated using ROC curves, and the results are as follows: Figure 7 As shown in the figure, the optimal plasma threshold for GPC1 to distinguish between individuals with ruptured aneurysms and those with unruptured aneurysms was 4.08 ng / ml, with an AUC of 0.831 (95% confidence interval: 0.678–0.792, P < 0.001), corresponding to a sensitivity of 89.7% and a specificity of 72.4%. The optimal plasma threshold for ECM2 to distinguish between individuals with ruptured aneurysms and those with unruptured aneurysms was 1605.87 ng / ml, with an AUC of 0.888 (95% confidence interval: 0.678–0.792, P < 0.001), corresponding to a sensitivity of 62.1% and a specificity of 100%. The optimal plasma threshold for PLEC to distinguish between individuals with ruptured aneurysms and those with unruptured aneurysms was 78.22 ng / ml, with an AUC of 0.853 (95% confidence interval: 0.678–0.792, P < 0.001), corresponding to a sensitivity of 89.7% and a specificity of 72.4%.

[0073] The ROC curve was used to evaluate the sensitivity and specificity of GPC1, ECM2 and PLEC in predicting aneurysm rupture. The results are shown in Figure 8 The AUC of GPC1 and PLEC in predicting aneurysm rupture was 0.925 (95% confidence interval: 0.825-0.978, P<0.001), and the corresponding sensitivity was 82.8% and the specificity was 86.2%. The AUC of PLEC and ECM2 in predicting aneurysm rupture was 0.904 (95% confidence interval: 0.797-0.965, P<0.001), and the corresponding sensitivity was 72.4% and the specificity was 96.6%. The AUC of GPC1 and ECM2 in predicting aneurysm rupture was 0.950 (95% confidence interval: 0.859-0.990, P<0.001), and the corresponding sensitivity was 82.8% and the specificity was 82.8%.

[0074] As Figures 5 to 8 compared with the results of using one or two of GPC1, ECM2 and PLEC to predict aneurysm rupture, the three factors were simultaneously included in the prediction model, which obviously showed a better area under the curve of prediction, and also improved the sensitivity and specificity of prediction.

[0075] The application also provides a kit for predicting intracranial aneurysm rupture using the biomarker combination. The detection reagent of the kit is composed of reagents for detecting the expression levels of biomarker combination factors GPC1, ECM2 and PLEC. The detection reagent as a key component of the kit can realize early warning of intracranial aneurysm rupture, and the above-mentioned prediction model can be used. When p>0.5, it is judged that the risk of intracranial aneurysm rupture is high, and treatment is recommended.

[0076] The application applies biomarker groups obtained based on high-throughput screening, including PLEC, GPC1 and ECM2, to establish a comprehensive prediction model for early warning of intracranial aneurysm rupture, reduce the treatment risk of patients and reduce the treatment pain of patients. The three methods of predicting aneurysm rupture, namely, PHASES scoring method, unruptured intracranial aneurysm treatment scoring method and the prediction model of the application, are used for evaluation on the set intracranial aneurysm patient population. The results show that the positive predictive value of PHASES scoring is 66.7%, the positive predictive value of unruptured intracranial aneurysm treatment scoring is 62.5%, and the positive predictive value of the prediction model is 95.0%. It is shown that the model is better than the existing international clinical scoring standards. The candidate markers are verified in the population, and the independent predictors of intracranial aneurysm rupture are determined by logistic regression model analysis. The logistic regression machine learning method is used for classification, 30% of the total samples are extracted as the test set, the remaining 70% of the samples are used as the training set for 10-fold cross-validation, and the sensitivity and accuracy of the model are evaluated by using the ROC curve. Finally, it is successful, has non-invasiveness and reliability, is easy to detect and calculate, and is convenient for wide application.

[0077] In conclusion, the purposes of the application are achieved.

Claims

1. The application of a combination of biomarkers in constructing a predictive model for intracranial aneurysm rupture, characterized in that: The biomarker combination consists of factors: GPC1, ECM2, and PLEC. A predictive model for intracranial aneurysm rupture is constructed based on these biomarkers.

2. The application of the prediction model according to claim 1, characterized in that: The formula for the prediction model is as follows: ln p / (1-p) = 1.7640*C GPC1 +0.0134*C ECM2 +0.1775*C PLEC +(-41.7000) C GPC1 C ECM2 and C PLEC The concentrations of GPC1, ECM2, and PLEC in the sample plasma are respectively assigned to specific measured values, and the probability p of aneurysm rupture risk in the sample is calculated.

3. A method for applying the prediction model according to claim 1 or 2, characterized in that: A predictive model for intracranial aneurysm rupture was constructed based on the combination of biomarkers, and a nomogram was established. The probability of rupture risk was evaluated by scoring the nomogram. Set up axes for plasma GPC1 concentration range, plasma PLEC concentration range, and plasma ECM2 concentration range, as well as a risk score axis, a total risk score axis, and a rupture risk probability axis. For each plasma concentration value of GPC1, PLEC, and ECM2, take the risk score value corresponding to the risk score axis, calculate the sum of each risk score value, find the corresponding total risk score value on the total risk score axis, and finally find the corresponding risk probability on the rupture risk probability axis.

Images