Construction method of accurate medication model for mycophenolic acid drugs, prediction device and electronic equipment

Abstract

The invention provides a construction method of a precise medication model for mycophenolic acid drugs, a prediction device and electronic equipment. Belongs to the cross technical field of precision medicine and pharmacology. The construction method of the precise medication model for the mycophenolic acid drugs comprises the following steps: S1, acquiring clinical data of a patient; s2, collecting blood samples at different time points after the patient takes the medicine, and detecting the blood concentration of the mycophenolic acid medicine; s3, the non-atrioventricular model is used as a basic model, multiple linear regression analysis is carried out, a precise medication model for the mycophenolic acid drugs is established, and a mathematical model of the precise medication model for the mycophenolic acid drugs is shown as a formula (1). By the adoption of the accurate medication model for the mycophenolic acid drugs, the blood sampling frequency for monitoring the blood concentration of the mycophenolic acid drugs can be reduced, the workload of medical workers is reduced, the pain of patients is relieved, and the compliance of the patients is improved.

Description

Technical Field

[0001] This invention relates to the interdisciplinary fields of precision medicine and pharmacology, and more specifically, to a method for constructing a precision medication model for mycophenolic acid drugs, a prediction device, and an electronic device. Background Technology

[0002] Organ transplantation is the most effective treatment for end-stage organ failure. The perioperative period of organ transplantation is a critical stage for postoperative recovery. The rational use of immunosuppressants after surgery is crucial to prevent graft rejection. Mycophenolic acid (MPA) drugs are commonly used immunosuppressants. They can reversibly inhibit inosine monophosphate dehydrogenase (IMPDH), a key enzyme in the de novo synthesis pathway of guanine nucleotides (GMP) in lymphocytes, thereby inhibiting lymphocyte proliferation and suppressing the immune response. After organ transplantation, MPA drugs are usually used in combination with other immunosuppressants such as tacrolimus and cyclosporine to achieve the best immunosuppressive effect.

[0003] Commonly used MPA drugs in clinical practice include mycophenolate mofetil capsules (MMF, such as Mycophenolate Mofetil, Cyclomethasone, and Mycophenolate Mofetil) and mycophenolate sodium enteric-coated tablets (EC-MPS, Mifu). When using MPA drugs, blood drug concentration monitoring is necessary to assess the in vivo exposure (AUC) of the MPA drug. 0-12h MPA has a narrow therapeutic window, and its efficacy and adverse reactions are closely related to blood drug concentration. The AUC of MPA in patients is also a factor. 0-12h Low-risk individuals will face a high risk of transplant rejection, MPA AUC 0-12h High levels of these drugs may also increase toxic side effects. Furthermore, due to differences in formulation and manufacturing processes, the pharmacokinetics of MPA drugs from different dosage forms and manufacturers vary significantly. Therefore, individualized therapeutic drug monitoring of MPA is crucial for optimizing clinical efficacy and reducing adverse reactions, and for controlling MPA AUC. 0-12h Maintaining a concentration of 30–60 mg / h / L can effectively reduce the incidence of rejection and decrease toxic side effects. Clinically, AUC is often calculated through intensive sampling, but this is difficult to implement due to the need for sampling at multiple time points, complex operation, poor patient compliance, and difficulty in implementation. Summary of the Invention

[0004] The purpose of this invention is to provide a precise medication model with fewer samples.

[0005] To achieve the above objectives, the first aspect of this invention provides a method for constructing a precision medication model for mycophenolic acid drugs, specifically including the following steps:

[0006] S1: Obtain patient clinical data;

[0007] S2: Collect blood samples at different time points after the patient takes the medication to detect the blood concentration of mycophenolic acid drugs;

[0008] S3: Based on the non-compartmental model, a precision medication model for mycophenolic acid drugs is established by performing multiple linear regression analysis. The mathematical model of the precision medication model for mycophenolic acid drugs is shown in formula (1):

[0009] (1)

[0010] Among them, AUC 0-12h The area under the curve for mycophenolic acid drugs from 0 to 12 hours is denoted as b, where b is a constant, n is the number of blood samples collected (1 ≤ n ≤ 6), and C is the total area under the curve. ti M represents the blood drug concentration obtained at time point ti. ti represents the coefficient of blood drug concentration in the model at time point ti.

[0011] This invention constructs a precision medication model for mycophenolic acid (MPA) drugs, as shown in formula (1), using multiple linear regression. This model reduces the blood collection burden on patients and improves patient compliance while ensuring prediction accuracy. Based on a non-compartmental model and experimental data, this model can dynamically adjust according to individual patient differences, accurately predicting the AUC of MPA drugs. 0-12h This optimizes drug dosage and reduces the risk of toxicity or rejection reactions caused by excessively high or low concentrations.

[0012] More preferably, the measured AUC of MPA 0-12h Phoenix® WinNonlin® version 8.35 software was used to calculate the non-compartmental model, and SPSS 25.0 was used for multiple linear regression analysis to establish the LSS model of MPA.

[0013] Preferably, in step S1, the clinical data includes the type of organ transplant, which is either a kidney transplant or a liver transplant.

[0014] The precision medication model provided by this invention is applicable to liver or kidney transplant patients. It establishes differentiated model parameters based on the differences in organ types between kidney and liver transplant patients. Using this model can fit the pharmacokinetic characteristics of patients with relevant organ transplants, thereby improving the accuracy of AUC prediction and its clinical applicability.

[0015] Preferably, if the organ transplant type is a kidney transplant, the clinical data also include whether postoperative graft function delay occurs.

[0016] Delayed graft function recovery (DGF) is a common early complication after organ transplantation, defined as the need for dialysis within one week post-transplantation or failure of serum creatinine to drop below 400 μmol / L within one week post-transplantation. This invention further refines the model's application scenarios by differentiating between DGF and non-DGF patients. Patients with DGF experience slower renal function recovery, which may affect MPA metabolic rate. The model provided by this invention can optimize sampling time points and coefficients for both DGF and non-DGF patients, enhancing the model's adaptability and predictive reliability in complex clinical situations.

[0017] Preferably, the clinical data also includes mycophenolic acid drug formulations, wherein the mycophenolic acid drug formulation is mycophenolate mofetil or mycophenolate sodium.

[0018] This invention establishes suitable models for two different dosage forms of MPA drugs: mycophenolate mofetil (such as cyclophosphamide, cyclophosphamide, and cyclophosphamide) and mycophenolate sodium (such as mifepristone). Because there are significant differences in absorption rate, bioavailability, and pharmacokinetic curves between different dosage forms (e.g., enteric-coated tablets versus capsules), the models provided by this invention can accurately reflect the pharmacokinetic characteristics of specific dosage forms, avoiding prediction biases caused by dosage form differences.

[0019] More preferably, based on the precision medication model shown in formula (1), if the patient is a non-DGF kidney transplant patient, when the mycophenolic acid drug is Mycophenolate mofetil, 3≤n≤6, n is most preferably 4, and the blood collection time points are 1h, 2h, 4h and 10h after medication.

[0020] More preferably, based on the precision medication model shown in formula (1), if the patient is a non-DGF kidney transplant patient, when the mycophenolic acid drug is Cyclopidogrel capsules, 2≤n≤6, and n is most preferably 2~4. When n is 2, the blood collection time points are 1h and 12h after medication, when n is 3, the blood collection time points are 1h, 2h and 12h after medication, and when n is 4, the blood collection time points are 0.5h, 1h, 2h and 12h after medication.

[0021] More preferably, based on the precision medication model shown in formula (1), if the patient is a non-DGF kidney transplant patient, when the mycophenolic acid drug is Changdianda, 3≤n≤6, and n is preferably 3 or 4. When n is 3, the blood collection time points are 1.5h, 4h and 6h after medication, respectively. When n is 4, the blood collection time points are 1.5h, 4h and 6h after medication, respectively.

[0022] More preferably, based on the precision medication model shown in formula (1), if the patient is a non-DGF kidney transplant patient, when the mycophenolic acid drug is Mifu, 4≤n≤6, n is most preferably 4, and when n is 4, the blood collection time points are 2h, 4h, 6h and 10h after medication.

[0023] More preferably, based on the precision medication model shown in formula (1), if the patient is a non-DGF kidney transplant patient, when the mycophenolic acid drug is cyclophosphamide dispersible tablets, 2≤n≤6, and n is most preferably 2~4. When n is 2, the blood collection time points are 0.5h and 2h after medication, respectively. When n is 3, the blood collection time points are 0.5h, 2h and 12h after medication, respectively. When n is 4, the blood collection time points are 0.5h, 2h, 8h and 12h after medication, respectively.

[0024] More preferably, based on the precision medication model shown in formula (1), if the patient is a DGF kidney transplant patient, when the mycophenolic acid drug is mycophenolate capsules (Cellulose, Changdianda, Cyclosporine), 2≤n≤4, n is most preferably 4, and when n is 4, the blood collection time points are 2h, 4h, 8h and 10h after medication.

[0025] More preferably, based on the precision medication model shown in formula (1), if the patient is a DGF kidney transplant patient, when the mycophenolic acid drug is mycophenolate sodium enteric-coated tablets (Mifu), n is 5 or 6, the most preferred n is 5, and the blood collection time points are 1.5h, 4h, 6h, 8h and 12h after medication.

[0026] More preferably, based on the precision medication model shown in formula (1), if the patient is a liver transplant patient, when the mycophenolic acid drug is Mycophenolate mofetil, 4≤n≤6, n is most preferably 4, and the blood collection time points are 1.5h, 2h, 4h and 12h after medication.

[0027] More preferably, based on the precision medication model shown in formula (1), if the patient is a liver transplant patient, when the mycophenolic acid drug is cyclophosphamide dispersible tablets, n is preferably 2, and the blood collection time points are 2h and 12h after medication.

[0028] More preferably, based on the precision medication model shown in formula (1), if the patient is a liver transplant patient, when the mycophenolic acid drug is mycophenolate capsules (Cellulose, Cyclosporine capsules), 4≤n≤6, n is most preferably 4, and the blood collection time points are 1.5h, 2h, 4h and 12h after medication.

[0029] Preferably, the construction method further includes S4: measuring model accuracy by prediction error and average prediction error, measuring model precision by root mean square error, and evaluating the consistency between measured values ​​and model predictions by Bland-Altman plot.

[0030] Preferably, the formula for calculating the prediction error is shown in formula (2):

[0031] (2)

[0032] Where PE is the prediction error, and AUC is the prediction error. predThe values ​​are model predictions and AUC is the measured value.

[0033] Preferably, the formula for the average prediction error is as shown in formula (3):

[0034] (3)

[0035] Where MPE is the average prediction error, n is the number of sampling points, i is the sample number, and PE is the average prediction error. i Let be the prediction error for the i-th sample.

[0036] Preferably, the formula for calculating the root mean square error is shown in formula (4):

[0037] (4)

[0038] Where RMSE is the root mean square error, n is the number of sampling points, i is the sample number, and AUC is the root mean square error. i Let AUC be the measured AUC value of the i-th sample. pred，i Predict the AUC value for the i-th sample using the model.

[0039] This invention validates the model from multiple dimensions using prediction error (PE), mean prediction error (MPE), root mean square error (RMSE), and Bland-Altman plots to ensure the model's accuracy and robustness in clinical practice. PE and MPE measure systematic bias, RMSE assesses random error, and the Bland-Altman plot visually demonstrates the consistency between measured and predicted values. These validation methods comprehensively ensure the model's reliability, providing clinicians with a credible basis for dose adjustment.

[0040] Furthermore, a second aspect of the present invention provides a dosing regimen prediction device for mycophenolic acid drugs, the dosing regimen prediction device for mycophenolic acid drugs comprising:

[0041] Data acquisition module: used to acquire clinical information, blood drug concentration data, and medication data of organ transplant recipients;

[0042] Medication regimen prediction module: used to determine the blood collection regimen of the patient based on the collected clinical information, medication data and blood drug concentration data, and to further predict the medication regimen based on the blood collection regimen. The blood collection regimen is used to guide the collection of blood samples from the patient, and the blood samples are used to determine the blood drug concentration of the corresponding mycophenolic acid drug.

[0043] Data storage module: Used to store patients' clinical information, blood drug concentration data, medication data, and blood collection and medication plans generated by the medication plan prediction module.

[0044] The medication regimen prediction device provided by this invention integrates data acquisition, prediction, and storage modules. It can automatically integrate patients' clinical information (such as organ type, presence of DGF, and drug dosage form) with real-time blood drug concentration data to quickly match the corresponding LSS blood collection protocol. Compared with traditional manual calculation or single model application, this device significantly shortens the medication regimen formulation time and reduces the operational complexity for clinicians. In addition, the medication regimen prediction module can dynamically select the optimal combination of blood collection time points based on the patient's characteristics, ensuring that the blood collection protocol is highly matched with the patient's individualized needs and reducing invalid sampling.

[0045] Furthermore, a third aspect of the present invention provides an electronic device, the electronic device comprising:

[0046] At least one processor, and

[0047] At least one computer-readable storage medium communicatively connected to the processor, the computer-readable storage medium storing program instructions executable by the processor, the program instructions being capable of performing the method for constructing a precision dosing model for mycophenolic acid drugs as described in the first aspect.

[0048] The precision medication model constructed by the method described in the first aspect, through the processor and storage medium of the electronic device, can achieve automated operation, meet the need for rapid generation of blood collection plans in complex clinical environments, and avoid errors from manual calculation.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] 1. Based on the Limited Sampling Strategy (LSS), this invention establishes a precision medication model for mycophenolic acid drugs, enabling clinicians to quickly determine the blood collection plan according to the patient's condition, greatly reducing the workload of blood collection;

[0051] 2. By using the precision medication model provided by this invention, a model of the relationship between AUC and blood drug concentration at each time point can be established by measuring the blood drug concentration at a few time points. This can reduce the number of samples required to calculate AUC and further greatly improve patient compliance.

[0052] 3. The precision medication model provided by this invention can help clinicians accurately predict patients' exposure to mycophenolic acids, which is beneficial for adjusting subsequent drug dosage and medication regimen. It can greatly reduce the incidence of adverse reactions to mycophenolic acids and the incidence of organ transplant rejection, and has significant economic and social benefits. Attached Figure Description

[0053] Figure 1 This is the consistency evaluation result between the measured and predicted values ​​of the precision medication model for non-DGF kidney transplant patients taking Mycophenolate mofetil in Example 1 of the present invention;

[0054] Figure 2 This is the consistency evaluation result between the measured and predicted values ​​of the precision medication model for non-DGF kidney transplant patients taking Cyclopidogrel capsules in Example 2 of the present invention;

[0055] Figure 3 This is the consistency evaluation result of the measured value and predicted value of the precision medication model for non-DGF kidney transplant patients taking Changdianda in Example 3 of the present invention;

[0056] Figure 4 This is the consistency evaluation result of the measured and predicted values ​​of the precision medication model for non-DGF kidney transplant patients taking Mifu in Example 4 of the present invention;

[0057] Figure 5 This is the consistency evaluation result of the measured value and predicted value of the precision medication model for non-DGF kidney transplant patients taking Cyclopine dispersible tablets in Example 5 of the present invention;

[0058] Figure 6 This is the consistency evaluation result of the measured and predicted values ​​of the precision medication model for DGF kidney transplant patients taking mycophenolate mofetil capsules (Xiaocic, Changdianda, and Cycopine capsules) in Example 6 of the present invention;

[0059] Figure 7 This is the consistency evaluation result between the measured and predicted values ​​of the precision medication model for DGF kidney transplant patients taking mycophenolate sodium enteric-coated tablets (Mifu) in Example 7 of the present invention;

[0060] Figure 8 This is the consistency evaluation result between the measured and predicted values ​​of the precision medication model for liver transplant patients taking Mycophenolate mofetil in Example 8 of the present invention;

[0061] Figure 9 This is the consistency evaluation result between the measured and predicted values ​​of the precision medication model for liver transplant patients taking Cyclopidogrel dispersible tablets in Example 9 of the present invention;

[0062] Figure 10 This is the consistency evaluation result of the measured and predicted values ​​of the precision medication model for liver transplant patients taking mycophenolate mofetil capsules (CellCyclophenol, Cyclophenol Capsules) in Example 10 of the present invention. Detailed Implementation

[0063] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described in detail below. It should be noted that the following embodiments are only used to illustrate the implementation methods and typical parameters of the present invention, and are not intended to limit the parameter range described in the present invention. Reasonable variations derived therefrom are still within the protection scope of the claims of the present invention.

[0064] It should be noted that the endpoints and any values ​​of the ranges disclosed herein are not limited to the precise ranges or values, and these ranges or values ​​should be understood to include values ​​close to these ranges or values. For numerical ranges, the endpoint values ​​of the various ranges, the endpoint values ​​of the various ranges and individual point values, and individual point values ​​can be combined with each other to obtain one or more new numerical ranges, which should be considered as specifically disclosed herein.

[0065] As described in the background section, in the prior art, in order to reduce the incidence of adverse reactions to MPA drugs and the incidence of rejection after organ transplantation, frequent blood draws are required to ensure the MPA AUC. 0-12h The concentration should be maintained at 30-60 mg / h. However, frequent blood sampling is complicated and patient compliance is poor. Therefore, the specific embodiments of the present invention provide a precision medication model for mycophenolic acid drugs.

[0066] This precision medicine model is constructed through the following steps:

[0067] S1: Obtain patient clinical data;

[0068] S2: Collect blood samples at different time points after the patient takes the medication to detect the blood concentration of mycophenolic acid drugs;

[0069] S3: Based on the non-compartmental model, a precision medication model for mycophenolic acid drugs is established by performing multiple linear regression analysis. The mathematical model of the precision medication model for mycophenolic acid drugs is shown in formula (1):

[0070] (1)

[0071] Among them, AUC 0-12h The area under the curve for mycophenolic acid drugs from 0 to 12 hours is denoted as b, where b is a constant, 1 ≤ n ≤ 6, and C is the area under the curve. ti M represents the blood drug concentration obtained at time point ti. ti represents the coefficient of blood drug concentration in the model at time point ti.

[0072] More specifically, the precision medication model provided by this invention is applicable to kidney transplant patients and liver transplant patients. If the organ transplant type is kidney transplant, the clinical data also includes whether postoperative graft function delay occurs.

[0073] More specifically, the precise medication model provided in the above embodiments targets mycophenolic acid drug formulations such as mycophenolate mofetil or mycophenolate sodium.

[0074] More specifically, the method for constructing a precision medication model provided by the above embodiments further includes S4: measuring the model accuracy by prediction error and average prediction error, measuring the model precision by root mean square error, and evaluating the consistency between measured values ​​and model predictions by Bland-Altman plot.

[0075] The formula for calculating the prediction error is shown in formula (2):

[0076] (2)

[0077] Where PE is the prediction error, and AUC is the prediction error. pred The values ​​are model predictions and AUC is the measured value.

[0078] The formula for the average prediction error is shown in formula (3):

[0079] (3)

[0080] Where MPE is the average prediction error, n is the number of sampling points, i is the sample number, and PE is the average prediction error. i Let be the prediction error for the i-th sample.

[0081] The formula for calculating the root mean square error is shown in formula (4):

[0082] (4)

[0083] Where RMSE is the root mean square error, n is the number of sampling points, i is the sample number, and AUC is the root mean square error. i Let AUC be the measured AUC value of the i-th sample. pred，i Predict the AUC value for the i-th sample using the model.

[0084] In some embodiments, the present invention also provides a blood collection protocol prediction device for mycophenolic acid drugs, the medication protocol prediction device for mycophenolic acid drugs comprising:

[0085] Data acquisition module: used to acquire clinical information, blood drug concentration data, and medication data of organ transplant recipients;

[0086] Medication regimen prediction module: used to determine the blood collection regimen of the patient based on the collected clinical information, medication data and blood drug concentration data, and to further predict the medication regimen based on the blood collection regimen. The blood collection regimen is used to guide the collection of blood samples from the patient, and the blood samples are used to determine the blood drug concentration of the corresponding mycophenolic acid drug.

[0087] Data storage module: Used to store patients' clinical information, blood drug concentration data, medication data, and blood collection and medication plans generated by the medication plan prediction module.

[0088] In some embodiments, the present invention also provides an electronic device, the electronic device comprising:

[0089] At least one processor, and

[0090] At least one computer-readable storage medium communicatively connected to the processor, the computer-readable storage medium storing program instructions executable by the processor, the program instructions being capable of executing the method for constructing a precision dosing model for mycophenolic acid drugs provided in the foregoing embodiments.

[0091] The technical solutions of the present invention are further described below through specific embodiments. Unless otherwise defined, all terms, symbols, and other scientific terms used herein are intended to have the same meaning as commonly understood by one of ordinary skill in the art. In some cases, terms with conventional meanings are limited herein for clarification or ease of reference, and such limitations should not be construed as indicating a significant difference from conventional understanding in the art. The technical methods described or referenced herein are generally well understood by those skilled in the art and have been employed by conventional methods. Unless otherwise stated, the use of commercially available kits, reagents, and instruments shall be performed according to the manufacturer's instructions and parameters.

[0092] Example 1

[0093] Construction of a precision dosing model for non-DGF kidney transplant patients taking mycophenolate mofetil

[0094] Non-DGF kidney transplant recipients of mycophenolate mofetil MPA AUC 0-12h Multiple linear regression equation and correlation r 2 See Table 1.

[0095] Table 1

[0096]

[0097] As shown in Table 1, the AUC predicted by a single blood drug concentration time point 0-12h The regression equation for MPA performs poorly, and the 2-point prediction of MPA AUC is weak. 0-12h The regression equation also showed unsatisfactory predictive performance, with predictions of MPA AUC at blood drug concentration time points from 3 to 6 points. 0-12h The multiple linear regression model showed good correlation.

[0098] Table 2 shows the validation results of the finite sampling method model in non-DGF recipients of the mycophenolate mofetil group in kidney transplantation.

[0099] Table 2

[0100]

[0101] Figure 1The Bland-Altman plot evaluated the MPA AUC. 0-12h The consistency between measured values ​​and model predictions. (From...) Figure 1 It is evident that as the number of sampling points increases, the model's prediction error gradually decreases, and the prediction accuracy improves. Figure 1 A (Model 1) in the data shows that the consistency of the two-point sampling method is poor, and the upper and lower limits are large. Figure 1 The upper and lower limits of B, C, and D (models 2, 3, and 4) decrease sequentially, especially in figure D (model 4), where the upper and lower limits are the smallest and the data points are the most concentrated, indicating that model 4 has the best predictive performance. Figure 1 E and F (models 5 and 6) further validated the positive impact of increasing sampling points on improving prediction accuracy. However, because models 5 and 6 involve more blood sampling points, resulting in excessive blood collection and increased patient discomfort, the 4-point model is more suitable for clinical needs.

[0102] Example 2

[0103] Construction of a precision medication model for non-DGF kidney transplant patients taking cyclophosphamide capsules

[0104] Non-DGF kidney transplant recipients of cyclophosphamide capsules MPA AUC 0-12h The equation and correlation r 2 See Table 3.

[0105] Table 3

[0106]

[0107] As shown in Table 3, the MPA AUC is predicted at a single blood drug concentration time point. 0-12h The regression equation showed poor predictive performance, and the prediction of MPA AUC at blood drug concentration time points from 2 to 6 was limited. 0-12h The multiple linear regression model showed good correlation.

[0108] Table 4 shows the validation results of the finite sampling method model in non-DGF recipients of the kidney transplantation cyproheptadine capsule group.

[0109] Table 4

[0110]

[0111] Figure 2 The Bland-Altman plot evaluated the MPA AUC. 0-12h The consistency between measured values ​​and model predictions. (From...) Figure 2 It is evident that as the number of sampling points increases, the model's prediction error gradually decreases, and the prediction accuracy improves. Figure 2 A (Model 1) in the data shows that the consistency of the two-point sampling method is poor, and the upper and lower limits are large. Figure 2The upper and lower limits of B, C, D, E, and F (models 2, 3, 4, 5, and 6) decrease sequentially, indicating a gradual increase in the predictive performance of the model. To reduce patient suffering, models with points 2-4 can be used for clinical needs.

[0112] Example 3

[0113] Construction of a precision medication model for non-DGF kidney transplant patients taking Changdianda

[0114] The MPA AUC of non-DGF recipients in the long-term kidney transplant group 0-12h Multiple linear regression equation and correlation r 2 See Table 5.

[0115] Table 5

[0116]

[0117] As shown in Table 5, the single plasma drug concentration time point and the two-point prediction of MPA AUC 0-12h The regression equation performed poorly, predicting MPA AUC at drug concentration time points from 3 to 6 points. 0-12h The multiple linear regression model showed good correlation.

[0118] Table 6 shows the validation results of the finite sampling method model for non-DGF recipients in the kidney transplantation Changdianda group.

[0119] Table 6

[0120]

[0121] Figure 3 The Bland-Altman plot evaluated the MPA AUC. 0-12h The consistency between measured values ​​and model predictions. (From...) Figure 3 As can be seen, the prediction error of the model gradually decreases and the prediction accuracy improves with the increase of sampling points. To reduce patient suffering, a 3-4 point model can be used for clinical needs.

[0122] Example 4

[0123] Construction of a precision drug administration model for non-DGF kidney transplant patients taking Mifu

[0124] MPA AUC in patients with non-DGF kidney transplant Mifu group 0-12h Multiple linear regression equation and correlation r 2 See Table 7.

[0125] Table 7

[0126]

[0127] As shown in Table 7, the time points of blood drug concentration 1-3 predict MPA AUC 0-12h The regression equation showed poor predictive performance, and the prediction of MPA AUC at blood drug concentration time points of 4-6 points was limited. 0-12h The multiple linear regression model showed good correlation.

[0128] Table 8 shows the validation results of the finite sampling method model in non-DGF recipients of the Mif group in the kidney transplantation group.

[0129] Table 8

[0130]

[0131] Figure 4 The Bland-Altman plot evaluated the MPA AUC. 0-12h The consistency between measured values ​​and model predictions is evident. As the number of sampling points increases, the model's prediction error gradually decreases, and its prediction accuracy improves. To reduce patient suffering, the 4-point model can be used for clinical needs.

[0132] Example 5

[0133] Construction of a precision medication model for non-DGF kidney transplant patients taking cyclophosphamide dispersible tablets

[0134] Kidney transplant recipients of cyclophosphamide dispersible tablets (Cyclophosphamide) in non-DGF recipients MPA AUC 0-12h Multiple linear regression equation and correlation r 2 See Table 9.

[0135] Table 9

[0136]

[0137] As shown in Table 9, the predicted MPA AUC at a single point in the blood drug concentration time point 0-12h The regression equation showed poor predictive performance, predicting MPA AUC at 2-4 time points of blood drug concentration. 0-12h The regression equation showed good predictive performance.

[0138] The results of the limited sampling method for non-DGF recipients of cyclophosphamide dispersible tablets in kidney transplantation are shown in Table 10.

[0139] Table 10

[0140]

[0141] The MPA AUC was analyzed using the Bland-Altman method. 0-12h The consistency evaluation results between the measured values ​​and the model predictions are shown in Figure 5 .Depend on Figure 5It is evident that when the number of sampling points is increased to 4, the model's prediction error is minimized and its accuracy is maximized. The Bland-Altman plot also shows the best consistency, with all points falling within the consistency range and having the smallest consistency range.

[0142] Example 6

[0143] Construction of a precision medication model for kidney transplant patients taking mycophenolate mofetil capsules based on DGF

[0144] MPA AUC in perioperative DGF recipients after kidney transplantation taking mycophenolate mofetil capsules (CellCyclophosphamide, Cyclophosphamide, Cyclophosphamide capsules) 0-12h Multiple linear regression equation and correlation r 2 See Table 11.

[0145] Table 11

[0146]

[0147] As shown in Table 11, the plasma drug concentration time points 1-4 predict the MPA AUC. 0-12h The regression equation showed good predictive performance.

[0148] Table 12 shows the results of DGF recipient limited sampling verification of mycophenolate mofetil capsules administered during the perioperative period of kidney transplantation.

[0149] Table 12

[0150]

[0151] The MPA AUC was analyzed using the Bland-Altman method. 0-12h The consistency evaluation results between the measured values ​​and the model predictions are shown in Figure 6 .Depend on Figure 6 It is evident that when the number of sampling points is increased to 4, the model's prediction error is minimized and its accuracy is maximized. The Bland-Altman plot also shows the best consistency, with all points falling within the consistency range and having the smallest consistency range.

[0152] Example 7

[0153] Construction of a precision medication model for mycophenolate sodium enteric-coated tablets in kidney transplant patients based on DGF

[0154] MPA AUC in perioperative DGF recipients after kidney transplantation taking mycophenolate sodium enteric-coated tablets (Mifu) 0-12h Multiple linear regression equation and correlation r 2 See Table 13.

[0155] Table 13

[0156]

[0157] As shown in Table 13, the time points of blood drug concentration 1-3 predict MPA AUC 0-12h The regression equation showed poor predictive performance, predicting MPA AUC at 4-5 time points of blood drug concentration. 0-12h The multiple linear regression model showed good correlation.

[0158] The results of the limited sampling method validation in the DGF recipient group of kidney transplant recipients who took mycophenolate sodium enteric-coated tablets (Mifu) are shown in Table 14.

[0159] Table 14

[0160]

[0161] The MPA AUC was analyzed using the Bland-Altman method. 0-12h The consistency evaluation results between the measured values ​​and the model predictions are shown in Figure 7 .Depend on Figure 7 It is evident that when the number of sampling points is increased to 5, the model's prediction error is minimized and its accuracy is maximized. The Bland-Altman plot also shows the best consistency, with all points falling within the consistency range and the consistency range being minimized.

[0162] Example 8

[0163] Precision medication model based on mycophenolate mofetil (MBPM) administration in liver transplant patients.

[0164] Liver transplant recipients of mycophenolate mofetil MPA AUC 0-12h Multiple linear regression equation and correlation r 2 See Table 15.

[0165] Table 15

[0166]

[0167] As shown in Table 15, the time points of blood drug concentration 1-3 predict MPA AUC 0-12h The predictive performance was poor, with 4-point (C1.5, C12, C4, C2), 5-point (C1.5, C12, C4, C2, C0.5), and 6-point (C1.5, C12, C4, C2, C0.5, C8) time-point predictions of MPA AUC. 0-12h The multiple linear regression model showed good correlation.

[0168] The results of the limited sampling method validation in the liver transplant mycophenolate mofetil group are shown in Table 16.

[0169] Table 16

[0170]

[0171] The MPA AUC was analyzed using the Bland-Altman method. 0-12h The consistency evaluation results between the measured values ​​and the model predictions are shown in Figure 8 .Depend on Figure 8 It is evident that as the number of sampling time points increases, the model's prediction error gradually decreases, while its prediction accuracy gradually improves. Furthermore, with the increase in sampling points, the Bland-Altman model... Figure 1 The better the consistency, the better. However, the 5-point and 6-point models require more blood collection points, leading to excessive blood collection and increased patient discomfort. Therefore, the 4-point model is more suitable for clinical needs.

[0172] Example 9

[0173] Construction of a precision medication model based on the use of cyclophosphamide dispersible tablets in liver transplant patients

[0174] MPA AUC in liver transplant patients in the cyclophosphamide dispersible tablet group 0-12h Multiple linear regression equation and correlation r 2 See Table 17.

[0175] Table 17

[0176]

[0177] As shown in Table 17, the predicted MPA AUC at a single point in the blood drug concentration time point 0-12h The prediction effect is better, and two points make the model's prediction correlation better.

[0178] The results of the limited sampling method validation for the liver transplantation cyproheptad dispersible tablets group are shown in Table 18.

[0179] Table 18

[0180]

[0181] The MPA AUC was analyzed using the Bland-Altman method. 0-12h The consistency evaluation results between the measured values ​​and the model predictions are shown in Figure 9 .Depend on Figure 9 It is evident that as the number of sampling time points increases, the model's prediction error gradually decreases, while its prediction accuracy gradually improves. Furthermore, with the increase in sampling points, the Bland-Altman model... Figure 1 The better the consistency, the better the prediction of MPA AUC. 0-12h ,

[0182] Two-point sampling is more difficult than single-point sampling, so the best sampling model is the single-point C2 sampling model.

[0183] Example 10

[0184] Construction of a precision medication model based on mycophenolate mofetil capsules in liver transplant patients

[0185] The MPA AUC of patients in the liver transplant mycophenolate mofetil capsule (CellCell, Cyclopramide capsule) group 0-12h Multiple linear regression equation and correlation r 2 See Table 19.

[0186] Table 19

[0187]

[0188] As shown in Table 19, the plasma drug concentration time points 1-3 predict the MPA AUC. 0-12h The predictive effect was poor; the prediction of MPA AUC at blood drug concentration time points between 4 and 6 points was relatively weak. 0-12h The multiple linear regression model showed good correlation.

[0189] The results of the limited sampling method validation of patients in the liver transplant mycophenolate mofetil capsule (CellCyclophenol, Cyclophenol Capsule) group are shown in Table 20.

[0190] Table 20

[0191]

[0192] The MPA AUC was analyzed using the Bland-Altman method. 0-12h The consistency evaluation results between the measured values ​​and the model predictions are shown in Figure 10 .Depend on Figure 10 It is evident that as the number of sampling time points increases, the model's prediction error gradually decreases, while its prediction accuracy gradually improves. Furthermore, with the increase in sampling points, the Bland-Altman model... Figure 1 The better the consistency, the better. However, the 5-point and 6-point models require more blood collection points, leading to excessive blood collection and increased patient discomfort. Therefore, the 4-point model is more suitable for clinical needs.

[0193] As can be seen from the above examples, based on the precision medication model shown in formula (1), if the patient is a non-DGF kidney transplant patient, when the mycophenolic acid drug is Mycophenolate mofetil, 3≤n≤6, and n is preferably 4, the blood collection time points are 1h, 2h, 4h, and 10h after medication. If the patient is a non-DGF kidney transplant patient, when the mycophenolic acid drug is Cyclopidogrel capsules, 2≤n≤6, and n is preferably 2~4, when n is 2, the blood collection time points are 1h and 12h after medication; when n is 3, the blood collection time points are 1h, 2h, and 12h after medication; and when n is 4, the blood collection time points are 0.5h, 1h, 2h, and 12h after medication. If the patient is a non-DGF kidney transplant recipient, and the mycophenolic acid drug is Changdianda, the optimal blood collection time is 3 ≤ n ≤ 6, with n preferably being 3 or 4. When n is 3, the blood collection time points are 1.5h, 4h, and 6h after drug administration; when n is 4, the blood collection time points are 1.5h, 4h, 6h, and 12h after drug administration. If the patient is a non-DGF kidney transplant recipient, and the optimal mycophenolic acid drug is Miflu, the optimal blood collection time is 4 ≤ n ≤ 6, with n preferably being 4. When n is 4, the blood collection time points are 2h, 4h, 6h, and 10h after drug administration. If the patient is a non-DGF kidney transplant recipient, when the mycophenolic acid drug is cyclophosphamide dispersible tablets, 2≤n≤6, with n preferably being 2~4, and most preferably 4. When n is 2, the blood collection time points are 0.5h and 2h after medication; when n is 3, the blood collection time points are 0.5h, 2h, and 12h after medication; and when n is 4, the blood collection time points are 0.5h, 2h, 8h, and 12h after medication. If the patient is a DGF kidney transplant recipient, when the mycophenolic acid drug is mycophenolate mofetil capsules (CellCell, Cyclophosphamide, Cyclophosphamide), 1≤n≤4, with n most preferably 4. When n is 4, the blood collection time points are 2h, 4h, 8h, and 10h after medication. If the patient is a DGF kidney transplant recipient, and the mycophenolic acid drug is mycophenolate sodium enteric-coated tablets (Mifu), n is 5 or 6, with n being the most preferred. Blood collection times are 1.5h, 4h, 6h, 8h, and 12h after medication. If the patient is a liver transplant recipient, and the mycophenolic acid drug is Mycophenolate mofetil, 4≤n≤6, with n being the most preferred. Blood collection times are 1.5h, 2h, 4h, and 12h after medication. If the patient is a liver transplant recipient, and the mycophenolic acid drug is Cyclomethasone dispersible tablets, n is the most preferred. Blood collection times are 2h and 12h after medication. If the patient is a liver transplant recipient, and the mycophenolic acid drug is mycophenolate mofetil capsules (Cercophenolate mofetil, Cyclomethasone capsules), 4≤n≤6, with n being the most preferred. Blood collection times are 1.5h, 2h, 4h, and 12h after medication.

[0194] While the disclosure is as stated above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of this disclosure, and all such changes and modifications will fall within the protection scope of this invention.

Claims

1. A method for constructing a precision medication model for mycophenolic acids, characterized in that, Includes the following steps: S1: Obtain patient's clinical data; S2: Collect blood samples at different time points after the patient takes the medication to detect the blood concentration of mycophenolic acid drugs; S3: Based on the non-compartmental model, a precision medication model for mycophenolic acid drugs is established by performing multiple linear regression analysis. The mathematical model of the precision medication model for mycophenolic acid drugs is shown in formula (1): （1） Among them, AUC 0-12h The area under the curve for mycophenolic acid drugs from 0 to 12 hours is denoted as b, where b is a constant, n is the number of sampling points, 1 ≤ n ≤ 6, and C is the area under the curve. ti M represents the blood drug concentration obtained at time point ti. ti represents the coefficient of blood drug concentration in the model at time point ti.

2. The construction method as described in claim 1, characterized in that, In step S1, the clinical data includes the type of organ transplant, which is either a kidney transplant or a liver transplant.

3. The construction method as described in claim 2, characterized in that, If the organ transplant is a kidney transplant, the clinical data also include whether postoperative graft function recovery was delayed.

4. The construction method as described in claim 1, characterized in that, The clinical data also includes mycophenolic acid drug formulations, namely mycophenolate mofetil or mycophenolate sodium.

5. The construction method as described in claim 1, characterized in that, The construction method further includes S4: measuring model accuracy by prediction error and average prediction error, measuring model precision by root mean square error, and evaluating the consistency between measured values ​​and model predictions by Bland-Altman plot.

6. The construction method as described in claim 5, characterized in that, The formula for calculating the prediction error is shown in formula (2): （2） Where PE is the prediction error, and AUC is the prediction error. pred The values ​​are model predictions and AUC is the measured value.

7. The construction method as described in claim 5, characterized in that, The formula for the average prediction error is shown in formula (3): （3） Where MPE is the average prediction error, n is the number of sampling points, i is the sample number, and PE is the average prediction error. i Let be the prediction error for the i-th sample.

8. The construction method as described in claim 5, characterized in that, The formula for calculating the root mean square error is shown in formula (4): （4） Where RMSE is the root mean square error, n is the number of sampling points, i is the sample number, and AUC is the root mean square error. i Let AUC be the measured AUC value of the i-th sample. pred，i Predict the AUC value for the i-th sample using the model.

9. A dosing regimen prediction device for mycophenolic acid drugs, characterized in that, The dosing regimen prediction device for mycophenolic acids includes: Data acquisition module: used to acquire clinical information, blood drug concentration data, and medication data of organ transplant recipients; Medication regimen prediction module: used to determine the blood collection regimen of the patient based on the collected clinical information, medication data and blood drug concentration data, and to further predict the medication regimen based on the blood collection regimen. The blood collection regimen is used to guide the collection of blood samples from the patient, and the blood samples are used to determine the blood drug concentration of the corresponding mycophenolic acid drug. Data storage module: Used to store patients' clinical information, blood drug concentration data, medication data, and blood collection and medication plans generated by the medication plan prediction module.

10. An electronic device, characterized in that, The electronic device includes: At least one processor, and At least one computer-readable storage medium communicatively connected to the processor, the computer-readable storage medium storing program instructions executable by the processor, the program instructions being capable of performing the method for constructing a precision dosing model for mycophenolic acid drugs as described in any one of claims 1 to 8.

Images