Construction method of aGVHD risk prediction model, prediction method, system, equipment and storage medium

Abstract

The invention specifically discloses a construction method of an aGVHD risk prediction model, a prediction method, a prediction system, equipment and a storage medium, and relates to the technical field of biological medicines. Monitoring the concentration of sCD26 / DPPIV and IL-1alpha in peripheral blood of a patient before transplantation is an effective index for predicting aGVHD occurrence, has important clinical significance for establishing a biological marker for early recognition of aGVHD occurrence, and has a guiding value for aGVHD targeted therapy.

Description

Technical Field

[0001] The present invention relates to the technical field of acute graft-versus-host disease risk prediction, and particularly to a method for constructing an aGVHD risk prediction model, a prediction method, a system, a device, and a storage medium. Background Art

[0002] Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is an important means to cure malignant and non-malignant blood diseases. Despite effective immunosuppressive prophylaxis, acute graft-versus-host disease (aGVHD) remains the main complication and cause of non-relapse death after allo-HSCT, with an incidence as high as 40-60%, and a mortality rate of 55% after its occurrence, greatly affecting the overall survival rate and quality of life of patients.

[0003] Currently, the clinical diagnosis of aGVHD mainly relies on the patient's clinical manifestations supplemented by invasive tissue biopsy. However, the complexity of clinical manifestations and sampling bias may lead to difficult diagnosis or treatment delay. CD26 / DPPIV is a serine protease on the cell surface that can selectively cleave polypeptide molecules with proline, alanine, or serine at the second position from the N-terminus. It is widely expressed on the surface of barrier tissues, stem cells, and various immune cells, is a co-stimulatory molecule in the T lymphocyte signaling pathway, and exists in the blood circulation in a soluble form. Currently, multiple studies have found that the concentration of soluble CD26 / DPPIV (sCD26 / DPPIV) in peripheral blood is related to the conditions of various solid tumors and can be used as an evaluation index for treatment efficacy. In addition, some studies have also found that the sCD26 / DPPIV concentration is related to myocardial infarction, and administering DPPIV enzyme inhibitors can be used for the treatment of myocardial infarction. Previous studies have shown that high expression of CD26 / DPPIV on the surface of CD4 + T lymphocytes indicates that the T cells are Th1 type, and its high expression is related to the production of Th1-type cytokines such as IFNγ, TNFα, IL-6, etc. The activation of T lymphocytes is crucial in the occurrence and development of aGVHD.

[0004] At present, the research on specific biological markers may be an effective means to achieve early prediction and identification of aGVHD. In existing research, multiple proteins, including soluble tumor necrosis factor receptor (sTNFR1), soluble human matrilysin (sST2), regenerating islet-derived protein 3α (REG3α), interleukin-8 (IL-8), interferon α (INF-α), T cell immunoglobulin and mucin domain molecule 3 (TIM3), interleukin-6 (IL-6), etc., can all predict the risk of aGVHD occurrence. At the same time, according to the combination of different proteins, a risk stratification for aGVHD is established. For example, the MAGIC risk stratification established using the levels of sTNFR1, Reg3α, and soluble ST2 is used to identify patients at high risk of aGVHD. However, existing research still cannot accurately predict the occurrence of aGVHD, and more convenient, effective, and clinically applicable rapid detection predictors are still urgently needed. sCD26 / DPPIV and IL-1α are closely related to T cell activation, but there is no relevant report on whether the concentrations of sCD26 / DPPIV and IL-1α in peripheral blood can be used to jointly predict the occurrence of aGVHD. Summary of the Invention

[0005] (1) Technical problems to be solved

[0006] In view of this, one of the main objectives of the present invention is to provide a method for constructing an aGVHD risk prediction model, as well as a prediction method, system, device, and storage medium. Monitoring the concentrations of sCD26 / DPPIV and IL-1α in the peripheral blood of patients before transplantation is an effective indicator for predicting the occurrence of aGVHD, which has important clinical significance for establishing biological markers for early identification of aGVHD occurrence and also has guiding value for targeted treatment of aGVHD.

[0007] (2) Technical solutions

[0008] To achieve the above objective, the present invention provides a method for constructing an aGVHD risk prediction model, and the construction method includes:

[0009] S1: Construct a data queue:

[0010] Select a number of target objects who have undergone hematopoietic stem cell transplantation, and obtain the concentrations of biomarkers at different time points before each target object undergoes hematopoietic stem cell transplantation; wherein, the biomarkers include CD26 and IL-1α;

[0011] A number of target objects are divided into an aGVHD occurrence group and an aGVHD non-occurrence group according to whether aGVHD occurs. Among them, for each target object in the aGVHD occurrence group, the biomarker concentrations before and after the onset of aGVHD are selected; based on the principle of obtaining biomarker concentrations of target objects in the aGVHD non-occurrence group that are similar to the time distribution of biomarker acquisition for all target objects in the aGVHD occurrence group, the biomarker concentrations of the target objects in the aGVHD non-occurrence group are obtained;

[0012] If a single target object has concentration data of multiple biomarkers, one concentration data of a biomarker is randomly selected for constructing a data queue;

[0013] S2: Correlation analysis between biomarkers and aGVHD:

[0014] The biomarkers in the data queue described in step S1: are subjected to correlation analysis with whether the target object has aGVHD, and it is determined that CD26 and IL-1α have statistical significance with whether the target object has aGVHD;

[0015] S3: Establishment and selection of an aGVHD risk prediction model:

[0016] Based on the CD26 and IL-1α concentrations in the data queue described in step S1:, a number of logistic regression models are established, and the logistic regression model with the highest AUC value is selected as the aGVHD risk prediction model;

[0017] The expression of the aGVHD risk prediction model is:

[0018] ln P / (1 - P) = a×IL-1α + b×CD26 + c;

[0019] Where P represents the probability of aGVHD occurrence;

[0020] IL-1α represents the IL-1α concentration;

[0021] CD26 represents the CD26 concentration;

[0022] a, b, and c represent the parameters obtained by training the severe aGVHD risk prediction model.

[0023] In one embodiment, the expression of the aGVHD risk prediction model is:

[0024] ln P / (1 - P) = 0.227×IL-1α + 0.006×CD26 - 8.329;

[0025] Where P represents the probability of aGVHD occurrence;

[0026] IL-1α represents the IL-1α concentration;

[0027] CD26 represents the CD26 concentration.

[0028] In one embodiment, the risk level of the target subject developing aGVHD is predicted based on the P value; wherein, the larger the P value, the higher the predicted risk level of the target subject developing aGVHD; conversely, the smaller the P value, the lower the predicted risk level of the target subject developing aGVHD.

[0029] In one embodiment, the P value is compared with a preset threshold to predict the risk level of the target subject developing aGVHD;

[0030] If the P value is greater than or equal to the preset threshold, it is predicted that the risk level of the target subject developing aGVHD is a high risk;

[0031] If the P value is less than the preset threshold, it is predicted that the risk level of the target subject developing aGVHD is a low risk.

[0032] In one embodiment, any number from 0.6 to 0.8 is selected as the preset threshold.

[0033] In one embodiment, the preset threshold is 0.609.

[0034] On the other hand, the present invention also provides an aGVHD risk prediction method, and the aGVHD risk prediction method includes the following steps:

[0035] T1: Obtain the CD26 and IL-1α concentrations of the target subject before receiving hematopoietic stem cell transplantation;

[0036] T2: Based on the CD26 and IL-1α concentrations obtained in step T1 and the aGVHD risk prediction model constructed by the construction method described in any one of claims 1 to 6, calculate the P value;

[0037] T3: Based on the P value calculated in step T2, predict the risk level of the target subject developing aGVHD after receiving hematopoietic stem cell transplantation.

[0038] On the other hand, the present invention also provides an aGVHD risk prediction system, and the aGVHD risk prediction system includes:

[0039] A data acquisition module for obtaining the CD26 and IL-1α concentrations of the target subject before receiving hematopoietic stem cell transplantation;

[0040] A calculation module for calculating the P value based on the CD26 and IL-1α concentrations obtained by the data acquisition module and the aGVHD risk prediction model constructed by the above construction method;

[0041] A prediction module, configured to predict the risk level of a target subject developing aGVHD after receiving a hematopoietic stem cell transplantation based on the P value calculated by the calculation module.

[0042] In another aspect, the present invention further provides an aGVHD risk prediction device, which includes:

[0043] At least one processor;

[0044] And a memory communicatively connected to the at least one processor;

[0045] Wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the above-mentioned aGVHD risk prediction method.

[0046] In another aspect, the present invention further provides a computer-readable storage medium storing computer instructions, and the computer instructions are used to cause the computer to execute the above-mentioned aGVHD risk prediction method.

[0047] In another aspect, the present invention further provides the use of reagents for detecting the concentrations of CD26 and IL-1α in the preparation of a product for determining whether a target subject will progress to aGVHD after receiving a hematopoietic stem cell transplantation, and the product includes one or a combination of a kit, a chip, a system, and a device.

[0048] In one embodiment, the target subject includes mammals and non-mammals. Examples of mammals include, but are not limited to, any member of the class Mammalia: humans, non-human primates such as chimpanzees and other apes and monkeys; farm animals such as cows, horses, sheep, goats, pigs; domestic animals such as rabbits, dogs, and cats; laboratory animals, including rodents such as rats, mice, and guinea pigs, etc. Examples of non-mammals include, but are not limited to, birds, fish, or other non-mammals, etc.

[0049] In one embodiment, the target subject is a human.

[0050] In one embodiment, the biomarker is derived from a test sample.

[0051] In one embodiment, the test sample includes any sample of body tissue, cells, or fluid, or any sample derived from the body, such as swabs, cleaning fluids, aspirates, or rinsing fluids, etc.

[0052] In one embodiment, the sample to be tested includes, but is not limited to, blood, serum, plasma, blood components, synovial fluid, urine, semen, saliva, feces, cerebrospinal fluid, gastric contents, vaginal secretions, mucus, tissue biopsy samples, tissue homogenates, bone marrow aspirates, bone homogenates, sputum, aspirates, wound exudates, swabs or swab rinses such as nasopharyngeal swabs, and other body fluids, etc.

[0053] In one embodiment, the sample to be tested is a blood sample or a blood-derived sample, such as serum or plasma or blood components.

[0054] In one embodiment, the blood sample or blood-derived sample is peripheral blood.

[0055] In one embodiment, the peripheral blood is fresh.

[0056] In one embodiment, the peripheral blood is frozen.

[0057] In one embodiment, the target subject is healthy.

[0058] In one embodiment, the target subject is non-healthy.

[0059] In one embodiment, the target subject has aGVHD.

[0060] In one embodiment, the target subject does not have aGVHD.

[0061] (III) Beneficial effects

[0062] The present invention provides a method for constructing an aGVHD risk prediction model, as well as a prediction method, system device, and storage medium.

[0063] Compared with the prior art, the following beneficial effects are achieved:

[0064] 1. High accuracy: The aGVHD risk prediction model provided by the present invention has a good performance in the classification of the overall sample, and can correctly predict the sample categories of about 77.8%. The high accuracy indicates that the model can effectively integrate the biological markers sCD26 / DPPIV and IL-1α concentrations, and make a relatively reliable judgment on the occurrence or non-occurrence of the target event.

[0065] 2. Strong sensitivity: Among the samples predicted by the aGVHD risk prediction model provided by the present invention as the target event occurs, the proportion of samples that actually truly have the target event is 85.7%. This index shows that the model has a strong correct prediction ability for samples that are actually positive examples.

[0066] 3. High specificity: In the validation set, there were 6 true positive cases, 3 false positive cases, 8 true negative cases, and 1 false negative case. Thus, the calculated specificity was True negative / (True negative + False positive) × 100% = 72.7%. This indicates that the model has good predictive ability for samples that are actually negative examples.

[0067] 4. First recognized the predictive role of peripheral blood sCD26 / DPPIV and IL-1α concentrations in the occurrence of aGVHD after allo-HSCT. Early monitoring of sCD26 / DPPIV and IL-1α concentrations can identify high-risk aGVHD patients as early as possible.

[0068] 5. First discovered that the combination of sCD26 / DPPIV and IL-1α concentrations can improve the prediction accuracy of aGVHD compared to single indicators.

[0069] (IV) Terms and Definitions

[0070] As used herein, the term "threshold" refers to a predetermined number used in an operation. For the determination of the threshold, different measurement techniques will yield different measurement results, and in other measurement techniques, this value may change, and thus an analog conversion may be required, which is within the scope of the skills of those skilled in the art. At the same time, it is easily understood that the thresholds for different biomarkers are different and need to be determined according to the actual detection situation, and the determination method is familiar to those skilled in the art.

[0071] As used herein, the term "test sample" refers to a composition obtained from or derived from a target subject, which contains cell entities and / or other molecular entities to be characterized and / or identified, for example, based on physical, biochemical, chemical, and / or physiological characteristics. The sample can be obtained from the subject's blood and other fluid samples and tissue samples of biological origin, such as biopsy tissue samples or tissue cultures or cells derived therefrom. The source of the tissue sample can be solid tissue, such as from fresh, frozen, and / or preserved organ or tissue samples, biopsy tissues, or aspirates; blood or any blood component; body fluid; cells at any time during an individual's pregnancy or development; or plasma.

[0072] As used herein, "sensitivity" refers to the proportion of subjects with a positive outcome who are correctly identified as positive (for example, correctly identifying those subjects who have the disease or medical condition for which they are being tested). For example, sensitivity can characterize the ability of a method to correctly identify the number of subjects in a group with TTP.

[0073] As used herein, the term “specificity” refers to the proportion of subjects with a negative outcome who are correctly identified as negative (e.g., those subjects correctly identified as not having the disease or medical condition being tested). For example, specificity can characterize the ability of a method to correctly identify the number of subjects in a group not having TTP.

[0074] As used herein, the term “ROC” or “ROC curve” refers to a receiver operating characteristic curve. An ROC curve can be a graphical representation of the performance of a binary classifier system. For any given method, an ROC curve can be generated by plotting sensitivity versus specificity at various threshold settings. Additionally, given at least one of three parameters (e.g., sensitivity, specificity, and threshold setting), an ROC curve can determine the value or expected value of any unknown parameter. A curve fit to the ROC curve can be used to determine the unknown parameter.

[0075] As used herein, the term “AUC” or “ROC-AUC” generally refers to the area under the receiver operating characteristic curve. This metric can provide a measure of the diagnostic utility of the method, taking into account the sensitivity and specificity of the method. Generally, ROC-AUC is in the range of 0.5 to 1.0, where a value close to 0.5 indicates that the method has limited diagnostic utility (e.g., lower sensitivity and / or specificity), while a value close to 1.0 indicates that the method has greater diagnostic utility (e.g., higher sensitivity and / or specificity). See, e.g., Pepe et al., “Limitations of the Odds Ratio in Gauging the Performance of a Diagnostic, Prognostic, or Screening Marker,” Am. J. Epidemiol 2004, 159(9):882-890, which is incorporated herein by reference in its entirety. According to Cook, “Use and Misuse of the Receiver Operating Characteristic Curve in Risk Prediction,” Circulation 2007, 11

[0076] 5:928-935, outlines other methods for characterizing diagnostic utility using likelihood functions, likelihood ratios, information theory, predictive values, calibration (including goodness of fit), and reclassification measures, which is incorporated herein by reference in its entirety.

[0077] As used herein, the term “positive predictive value” or “PPV” refers to the probability that a subject has a positive outcome given that the subject has a positive test result.

[0078] As used herein, the term "negative predictive value" or "NPV" refers to the probability that a subject has a negative outcome assuming the subject has a negative test result. Brief Description of the Drawings

[0079] 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 for the description of the embodiments or the prior art. 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.

[0080] Figure 1 It is a ROC curve graph of sCD26 / DPPIV and IL-1α concentrations and combined prediction in the training set for the occurrence of grade II-IV aGVHD in patients.

[0081] Figure 2 It is a ROC curve graph of sCD26 / DPPIV and IL-1α concentrations and combined prediction in the validation set for the occurrence of grade II-IV aGVHD in patients. Detailed Embodiments

[0082] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0083] As used herein, "comprising", "having", or "including" includes "containing", "consisting essentially of...", "consisting substantially of...", and "consisting of..."; "consisting essentially of...", "consisting substantially of...", and "consisting of..." are subordinate concepts of "comprising", "having", or "including".

[0084] The experimental methods used in the following embodiments are all conventional methods unless otherwise specified. The reagents, methods, and equipment used are all conventional reagents, methods, and equipment in the technical field unless otherwise specified.

[0085] Example 1

[0086] Detection of CD26 / DPPIV concentration in patients' peripheral blood before reinfusion:

[0087] (1) Collect non-anticoagulated tube specimens of peripheral blood from patients who first received unrelated umbilical cord blood transplantation in the Department of Hematology, The First Affiliated Hospital of University of Science and Technology of China (Anhui Provincial Hospital) before reinfusion.

[0088] (2) Centrifuge at 1200 g, 4 °C for 5 min, collect 100 μl of serum, and store it in a -80 °C refrigerator.

[0089] (3) According to the instructions, use the DPPIV-Glo protease assay kit (G8350, Promega, Madison, WI) to detect the concentration of CD26 / DPPIV.

[0090] (4) Dissolve all serum specimens to be tested and purified DPPIV enzyme (9168-SE, R&D System s) at room temperature, and dilute them with 10 mM Tris-HCl (pH 8.0) at a ratio of 1:1000.

[0091] (5) Thaw the thawing buffer and lyophilized luciferin detection reagent in the assay kit to room temperature.

[0092] (6) Add 25 μl of ultrapure water to the substrate bottle to resuspend the substrate. Briefly mix by vortexing to form a 10 mmol / L stock solution.

[0093] (7) Add 10 ml of buffer to the luciferin detection reagent in the amber bottle to form mixture A.

[0094] (8) Add 20 μl of substrate to mixture A, and mix by rotating or inverting the contents to obtain a homogeneous mixture B.

[0095] (9) Add 25 μl of mixture B, 25 μl of 10 mM Tris-HCl (pH 8.0), and 25 μl of diluted serum sample or purified DPPIV enzyme to a 384-well plate in a 1:1 ratio.

[0096] (10) Gently mix the contents in the wells with a plate shaker at 300 - 500 rpm for 30 seconds. Incubate at room temperature for 30 min to 3 h.

[0097] (11) Use a microplate reader to record the luminescence, and draw a standard curve to quantify the DPPIV concentration in the serum.

[0098] Example 2

[0099] Detection of IL-1α concentration in peripheral blood before autologous blood transfusion of patients:

[0100] Peripheral blood non-anticoagulant tube specimens were collected from patients who first received unrelated umbilical cord blood transplantation in the Hematology Department of the First Affiliated Hospital of the University of Science and Technology of China (Anhui Provincial Hospital) before reinfusion. According to the above steps, serum was collected for CBA detection. The detection used the Bio-Plex Pro Human Cytokine Screening Panel (BIO-RAD, 12007283) kit.

[0101] Before the formal experiment, a pre-experiment was carried out. After determining the appropriate dilution ratio of the specimens, the formal experiment was carried out. The general steps of the experiment are as follows:

[0102] (1) According to the kit instructions, transfer the lyophilized standard microspheres to a 15 ml centrifuge tube, then add the specified volume of Assay Diluent to dissolve the standard, gently pipette to mix evenly, place at room temperature for at least 15 minutes, and label this tube as the highest concentration;

[0103] (2) Dilute the standard with Assay Diluent, and the dilution ratios are 1:2, 1:4, 1:8, 1:16, 1:32, 1:64, 1:128, and 1:256 respectively. Note to pipette evenly; Take another centrifuge tube and add only Assay Diluent as the negative control tube;

[0104] (3) According to the total number of test samples, standards, and control tubes, take an appropriate amount of mixed capture microspheres, add 0.5 ml of washing agent, centrifuge at 200×g at room temperature for 5 minutes, discard the supernatant, add an appropriate volume of dilution solution, resuspend the microspheres, and incubate at room temperature for 15 - 30 minutes;

[0105] (4) Prepare an appropriate amount of PE detection antibody according to the total number of test samples, standards, and control tubes, and store it in the dark at 4℃;

[0106] (5) Vortex the mixed capture microspheres thoroughly. According to the instructions, gently mix the specified volume of the test sample, standard, and negative control tube with the prepared mixed capture microspheres, and incubate at room temperature in the dark for 1 hour;

[0107] (6) Add the PE detection antibody, gently mix, and incubate at room temperature in the dark for 2 hours;

[0108] (7) Add 1 ml of washing agent to each tube, centrifuge at 200×g at room temperature for 5 minutes, discard the supernatant, add 200 μl of washing solution to each tube to resuspend the cells, and immediately perform on-machine detection;

[0109] (8) Record the data, use the FCAP Array Software (v3.0, 652099, BD Biosciences) to draw the standard curve, and calculate the concentration of IL-1α in the sample according to the standard curve.

[0110] Example 3

[0111] Statistical analysis:

[0112] The collected samples of allo-HSCT patients were divided into a training set and a validation set. In the training set, the correlation between the concentrations of CD26 / DPPIV and IL-1α in the peripheral blood before patient reinfusion and the occurrence of aGVHD was analyzed. According to the aGVHD situation, univariate and multivariate regression analyses of the concentrations of CD26 / DPPIV and IL-1α were performed to construct a prediction model, and the area under the curve (AUC) in the receiver operating characteristic (ROC) curve was used to evaluate the specificity of the prediction model, and the model was verified in the validation set. All data were expressed as mean ± standard deviation. Unpaired Student's two-tailed t-test was used to compare the means of the two groups, and statistical analysis was performed using R software version 4.2.2 and GraphPad Prism 9.0 software.

[0113] 1. In the training set, the concentrations of CD26 / DPPIV and IL-1α in the peripheral blood before patient reinfusion were positively correlated with the occurrence of aGVHD

[0114] A total of 59 specimens were collected from patients who received allogeneic hematopoietic stem cell transplantation for the first time before reinfusion, and the concentrations of sCD26 / DPPIV and IL-1α in the serum were detected. The data of 41 patients (70%) were selected as the training set, and the data of 18 patients (30%) were selected as the validation set for the correlation analysis of the concentrations of sCD26 / DPPIV and IL-1α and the occurrence of aGVHD in patients (in most statistical analyses of model construction, 70% of the total sample size is default selected as the training set, and 30% as the validation set). In the training set, for 41 patients who received allogeneic peripheral blood stem cell transplantation (allo-PBSCT) for the first time, the concentrations of sCD26 / DPPIV and IL-1α in the peripheral blood before reinfusion were detected, and the medical records of the patients were collected. According to the occurrence of aGVHD after transplantation, the patients were grouped, and the concentration differences before reinfusion between the two groups of patients were compared. The detection showed that the concentrations of sCD26 / DPPIV and IL-1α in the peripheral blood before transplantation in patients who developed grade II-IV aGVHD after transplantation were significantly higher than those in patients who did not develop or developed grade I aGVHD (sCD26 / DPPIV: 1154.0 ng / ml vs. 830.1 ng / ml, P = 0.001; IL-1α: 14.88 pg / ml vs. 9.13 pg / ml, P = 0.002). According to the concentrations of sCD26 / DPPIV and IL-1α in patients and the occurrence of grade II-IV aGVHD, ROC curve analysis was performed, and the areas under the curves were 0.797 (95% CI,

[0115] 0.654 to 0.939, P = 0.001) and 0.721 (95% CI, 0.552 to 0.890, P = 0.017), suggesting a moderate correlation between the concentration of sCD26 / DPPIV in peripheral blood before reinfusion and the occurrence of grade II-IV aGVHD after transplantation.

[0116] The concentrations of sCD26 / DPPIV and IL-1α were included for univariate and multivariate logistic regression analysis of grade II-IV aGVHD. In univariate analysis, both the concentrations of sCD26 / DPPIV and IL-1α were independent risk factors for aGVHD (sCD26 / DPPIV, Odds ratio: 1.004, 95% CI, 1.001 to 1.007, P = 0.006; IL-1α, Odds ratio:

[0117] 1.270, 95% CI, 1.040 to 1.560, P = 0.019). In multivariate analysis, both the concentrations of sCD26 / DPPI V and IL-1α were independent risk factors (sCD26 / DPPIV, Odds ratio:

[0118] 1.006, 95% CI, 1.002 to 1.011, P = 0.016; IL-1α, Odds ratio:

[0119] 1.255, 95% CI, 1.028 to 1.712, P = 0.044). According to the regression analysis results, a prediction model was constructed:

[0120] lnP / (1 - P) = 0.227×IL-1α + 0.006×sCD26 / DPPIV - 8.329

[0121] where IL-1α is the concentration (pg / ml) in peripheral blood before reinfusion of the patient, sCD26 / DPPIV is the concentration (ng / ml) in peripheral blood before reinfusion of the patient, and P is the probability of the patient developing grade II-IV aGVHD after transplantation. The ROC curve of the model was plotted, and the area under the curve was 0.900 (95% CI, 0.780 to 1.000, P < 0.001), which was significantly higher than the area under the curve for predicting the occurrence of grade II-IV aGVHD by sCD26 or IL-1α (P = 0.049 and 0.015, Figure 1 ). According to the combined prediction ROC curve, the optimal threshold was 0.609, and the P value was calculated according to the equation. If it was greater than 0.609, the risk of occurrence was high.

[0122] 2. In the validation set, the concentration of CD26 / DPPIV combined with IL-1α before reinfusion of the patient was significantly correlated with the occurrence of aGVH D;

[0123] In the validation set, the correlations between the concentrations of sCD26 / DPPIV and IL-1α in peripheral blood before reinfusion and aGVHD after transplantation were analyzed in 18 patients undergoing allo-PBSCT. The detection results showed that, for patients who developed grade II-IV aGVHD after transplantation, the concentrations of sCD26 / DPPIV and IL-1α in peripheral blood before transplantation were higher than those in patients who did not develop or developed grade I aGVHD (sCD26 / DPPIV: 1185.0 ng / ml vs. 1036 ng / ml, P = 0.204; IL-1α: 14.80 pg / ml vs. 9.32 pg / ml, P = 0.050). ROC curve analysis was performed based on the concentrations of sCD26 / DPPIV and IL-1α in patients and the occurrence of grade II-IV aGVHD. The areas under the curves were 0.688 (95% CI, 0.435 - 0.942, P = 0.189) and 0.753 (95% CI, 0.508 - 0.998, P = 0.077) respectively. Using the concentrations of sCD26 / DPPIV or IL-1α, the occurrence of grade II-IV aGVHD could not be accurately predicted in the validation set. When the concentrations of sCD26 / DPPIV and IL-1α were used in combination to predict the occurrence of aGVHD, in the ROC analysis, the area under the curve was 0.883 (95% CI, 0.713 - 1.000, P = 0.008), which was significantly higher than the areas under the curves for predicting aGVHD using individual concentrations ( Figure 2 ).

[0124] The risk prediction model obtained from the training set was used for validation in the validation set. The Logistics regression prediction model based on biological markers constructed in the present invention was comprehensively and rigorously evaluated for performance in the validation set, aiming to test the prediction ability and accuracy of the model for the occurrence of target events.

[0125] 1) Accuracy

[0126] The risk prediction model constructed based on the training set is ln P / (1 - P) = 0.227×IL-1α + 0.006×sCD26 / DPPIV - 8.329, where IL-1α is the concentration (pg / ml) in the peripheral blood of the patient before reinfusion, and sCD26 / DPPIV is the concentration (ng / ml) in the peripheral blood of the patient before reinfusion. Calculate the probability (P) of grade II-IV aGVHD occurring after transplantation in 18 patients in the validation set. Group the patients according to the cut-off value of 0.609, where those greater than 0.609 are at high risk. Evaluate the sensitivity and specificity of the model based on the actual occurrence of aGVHD in these patients after transplantation. In the validation set, there are 6 true positive cases, 3 false positive cases, 8 true negative cases, and 1 false negative case. Thus, the accuracy is calculated as (True positive + True negative) / (True positive + False positive + True negative + False negative)×100% = 77.8%.

[0127] This result indicates that the risk prediction model provided by the present invention has good performance in the classification of the overall sample and can correctly predict the sample categories of approximately 77.8%. The relatively high accuracy shows that the model can effectively integrate the biological markers sCD26 / DPPIV and the concentration of IL-1α and make a relatively reliable judgment on the occurrence of the target event.

[0128] 2) Sensitivity

[0129] In the validation set, there are 6 true positive cases, 3 false positive cases, 8 true negative cases, and 1 false negative case. Thus, the sensitivity is calculated as True positive / (True positive + False negative)×100% = 85.7%.

[0130] This means that among the samples predicted by the model as the occurrence of the target event, the proportion of samples actually having the target event is 85.7%. This index indicates that the model has a strong ability to correctly predict samples that are actually positive cases.

[0131] 3) Specificity

[0132] In the validation set, there are 6 true positive cases, 3 false positive cases, 8 true negative cases, and 1 false negative case. Thus, the specificity is calculated as true negative / (true negative + false positive) × 100% = 72.7%.

[0133] This indicates that the model has good correct prediction ability for samples that are actually negative examples.

[0134] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0135] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for constructing an aGVHD risk prediction model, characterized in that, The construction method includes the following: S1: Construct a data queue: Select a number of target subjects who have received hematopoietic stem cell transplantation, and obtain the concentrations of biomarkers at different time points before hematopoietic stem cell transplantation for each target subject; wherein, the biomarkers include CD26 and IL-1α; Divide the number of target subjects into an aGVHD occurrence group and an aGVHD non-occurrence group according to whether aGVHD occurs. Among them, for each target subject in the aGVHD occurrence group, select the biomarker concentrations before and after the onset of aGVHD; based on the principle of similar time distribution of biomarker acquisition for all target subjects in the aGVHD occurrence group, obtain the biomarker concentrations of the target subjects in the aGVHD non-occurrence group; If a single target subject has concentration data of multiple biomarkers, randomly select the concentration data of one of the biomarkers for constructing the data queue; S2: Correlation analysis of biomarkers and aGVHD: Perform a correlation analysis on the biomarkers in the data queue in step S1 and whether aGVHD occurs in the target subjects, and determine whether CD26 and IL-1α have statistical significance with whether aGVHD occurs in the target subjects; S3: Establishment and selection of an aGVHD risk prediction model: Based on the CD26 and IL-1α concentrations in the data queue in step S1, establish a number of logistic regression models, and select the logistic regression model with the highest AUC value as the aGVHD risk prediction model; The expression of the aGVHD risk prediction model is: ln P / (1 - P) = a × IL-1α + b × CD26 + c; Wherein, P represents the probability of aGVHD occurrence; IL-1α represents the concentration of IL-1α; CD26 represents the concentration of CD26; a, b, and c represent the parameters obtained by training the aGVHD risk prediction model.

2. The construction method according to claim 1, characterized in that The expression of the aGVHD risk prediction model is: ln P / (1 - P) = 0.227 × IL-1α + 0.006 × CD26 - 8.329; Wherein, P represents the probability of aGVHD occurrence; IL-1α represents the concentration of IL-1α; CD26 represents the concentration of CD26.

3. The construction method according to claim 1 or 2, characterized in that, Predict the risk degree of aGVHD occurrence in the target subject based on the P value; wherein, the larger the P value, the higher the predicted risk degree of aGVHD occurrence in the target subject; conversely, the smaller the P value, the lower the predicted risk degree of aGVHD occurrence in the target subject.

4. The construction method according to claim 3, wherein Compare the P value with a preset threshold to predict the risk degree of aGVHD occurrence in the target subject; If the P value is greater than or equal to the preset threshold, predict that the risk degree of aGVHD occurrence in the target subject is high risk; If the P value is less than the preset threshold, predict that the risk degree of aGVHD occurrence in the target subject is low risk.

5. The construction method according to claim 4, characterized in that The preset threshold selects any number from 0.6 to 0.

8.

6. A method for predicting the risk of aGVHD, characterized in that, The aGVHD risk prediction method includes the following steps: T1: Obtain the CD26 and IL-1α concentrations of the target subject before receiving hematopoietic stem cell transplantation; T2: Calculate the P value based on the CD26 and IL-1α concentrations obtained in step T1 and the aGVHD risk prediction model constructed by the construction method described in any one of claims 1 to 5. T3: Predict the risk degree of the target subject developing aGVHD after receiving hematopoietic stem cell transplantation based on the P value calculated in step T2.

7. A aGVHD risk prediction system, characterized in that, The aGVHD risk prediction system includes: A data acquisition module for acquiring the CD26 and IL-1α concentrations of the target subject before receiving hematopoietic stem cell transplantation. A calculation module for calculating the P value based on the CD26 and IL-1α concentrations acquired by the data acquisition module and the aGVHD risk prediction model constructed by the construction method described in any one of claims 1 to 5. A prediction module for predicting the risk degree of the target subject developing aGVHD after receiving hematopoietic stem cell transplantation based on the P value calculated by the calculation module.

8. An aGVHD risk prediction device, characterized in that, The aGVHD risk prediction device includes: At least one processor; And a memory communicatively connected to the at least one processor; Wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the aGVHD risk prediction method described in claim 6.

9. A computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to execute the aGVHD risk prediction method described in claim 6.

10. Use of a reagent for detecting the concentrations of CD26 and IL-1α in the preparation of a product for determining whether a target subject will progress to aGVHD after receiving hematopoietic stem cell transplantation, characterized in that, The product includes one or a combination of a kit, a chip, a system, a device, etc.

Images