Oxcarbazepine quantitative pharmacological model based on genetic polymorphism and construction method thereof

Abstract

The application discloses a genetic polymorphism-based oxcarbazepine quantitative pharmacological model and a construction method thereof, and belongs to the field of pharmacokinetics. The model construction method is characterized by the following steps: collecting steady-state blood drug concentration data and clinical data of epilepsy children patients, adopting a nonlinear mixed effect model method to establish a population pharmacokinetic model of an active metabolite of oxcarbazepine, and quantitatively investigating the influence of physiological factors, combined medication and genetic factors on pharmacokinetic characteristic parameters of the drug; and finally, the final model is internally verified and evaluated by using a goodness-of-fit diagram, a bootstrap method and an intuitive prediction test. The quantitative pharmacological model constructed by the application has stable and accurate prediction performance; the body weight and the genetic polymorphism ABCC2rs2273697 are determined as effective covariates influencing MHD CL / F; and an oxcarbazepine dose optimization scheme is established based on the body weight and the genotype of the patients, so that the treatment effect of the oxcarbazepine is improved, and the occurrence of adverse reactions is avoided.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of pharmacokinetics, and particularly relates to an oxcarbazepine quantitative pharmacological model based on genetic polymorphism and a construction method thereof. BACKGROUND

[0002] Epilepsy is one of the common chronic nervous system diseases, and its characteristic is to cause repeated seizures of epilepsy due to abnormal discharge of brain neurons. Children are a high-risk group of epilepsy, and 4 to 9 children per 1000 children suffer from epilepsy, which seriously affects the quality of life of countless children's families.

[0003] Oxcarbazepine is currently recommended by the International League Against Epilepsy (ILAE) as a first-line antiepileptic drug for children, and has a wide effect on various seizures. After oral administration, oxcarbazine is rapidly absorbed in the body and rapidly and almost completely degraded into the pharmacologically active metabolite 10-hydroxy carbamazepine (MHD), and the effective blood drug concentration range is 3-35 mg / L. Since children are in the growth and development period, the physiological function is not mature, and the metabolic enzyme system and organ function in the body are not developed; and MHD has the characteristics of large individual metabolic difference and nonlinear protein binding, which leads to large individual difference in blood drug concentration under the same dose, and further causes treatment failure or serious adverse reactions.

[0004] Compared with traditional pharmacokinetic research, population pharmacokinetic (PPK) research can accurately estimate the error caused by individual and intra-individual variation using sparse drug concentration data, and a small amount of blood sampling points are also conducive to the research on children's population, which is an effective method for designing individualized treatment plan. Compared with empirical dosing, model-guided precision dosing (MIPD) is a new method for developing a dosing regimen based on patient's physiological, pathological, genetic, disease and other characteristics, which can improve the safety and effectiveness of drug treatment. SUMMARY

[0005] The present application provides an oxcarbazepine quantitative pharmacological model based on genetic polymorphism and a construction method thereof. It aims to solve the problem that children with epilepsy have not fully developed metabolic enzyme systems and organ functions in the body, resulting in large differences in blood drug concentration among different individuals after taking oxcarbazepine, which causes treatment failure or serious adverse reactions. The present application adopts a nonlinear mixed effect model method, combines classical pharmacokinetic theory and statistical principles to establish a population pharmacokinetic model of the active metabolite MHD of oxcarbazepine, and quantitatively investigates the influence of physiological factors, combined medication and genetic factors on pharmacokinetic characteristic parameters. The final model uses goodness-of-fit plot, bootstrap method and visual predictive check for internal validation and evaluation. In the application of the model, the Monte Carlo simulation method is used to predict the MHD trough concentration under different dosing regimens, and dose recommendations are proposed for different subgroups of patients, so as to realize safe, effective and individualized drug use.

[0006] The technical scheme of the present application is implemented as follows:

[0007] A method for constructing an oxcarbazepine quantitative pharmacological model based on genetic polymorphism, comprising the following steps:

[0008] (1) Designing a research plan: adopting population pharmacokinetic analysis, formulating inclusion and exclusion criteria, and selecting genetic polymorphism sites related to drug transporters, drug metabolism and oxcarbazepine efficacy;

[0009] (2) Collecting data: collecting population statistics, biochemical indicators, medication, combined medication, blood concentration detection data of oxcarbazepine active metabolite MHD, and genotyping of genetic polymorphism sites;

[0010] (3) Basic model: using a one-compartment model (ADVAN2 TRANS2) with first-order absorption and elimination for modeling; an exponential model is used for fitting the inter-individual variation model, and a mixed type is used for fitting the residual variation model;

[0011] (4) Fixed effect model: using the forward inclusion method / reverse elimination method to establish a covariate model; screening the fixed effect parameters that significantly affect the pharmacokinetics of oxcarbazepine active metabolite MHD from various influencing factors, and plotting the covariates to be investigated and the individual clearance rate estimated by the empirical Bayes method; obtaining the final model; body weight and genetic polymorphism ABCC2 rs2273697 are retained in the final model;

[0012] (5) Model evaluation: the stability and prediction performance of the model are evaluated by graphical and statistical methods; after comprehensive evaluation, it is proved that the final model is stable, reliable and has high prediction performance;

[0013] (6) Dose optimization: according to the included covariates, virtual patients are divided into different subgroups, and based on the final model, Monte Carlo simulation method is used to predict the distribution of steady-state blood drug concentration under different dosing regimens, and the model construction is completed, which provides a basis for individualized and rational drug use for children with epilepsy.

[0014] Preferably, in step (1):

[0015] The inclusion and exclusion criteria include inclusion criteria and exclusion criteria; the inclusion criteria are: 1) children under 12 years old with seizure types in accordance with the International League Against Epilepsy (ILAE); 2) strictly following the doctor's advice according to the diagnosis; 3) during the study, no special treatment such as dialysis or diuresis that significantly affects drug elimination; the exclusion criteria are: 1) missing key research data, including unknown dose, no recent weight record, abnormal weight change; 2) unclear blood drug concentration monitoring information or sampling time does not meet the requirements; 3) significant abnormalities in liver and kidney function during medication.

[0016] The genetic polymorphism sites include ABCB1 rs1128503, ABCB1 rs3789243, ABCB1 rs1045642, ABCC2 rs2273697, ABCC2 rs717620, ABCC2 rs3740066, ADORA2Ars2298383, AS3MT rs7085104, IL1B rs16944, MTHFR rs1801131, MTHFR rs1801133, SCN1A rs6730344, SCN1A rs2298771, SCN1A rs10167228, SCN1Ars3812718, SCN1A rs6732655, SCN2A rs2304016, SCN2A rs17183814, UGT2B7 rs7668258, UGT2B7 rs7668282, UGT2B7 rs12233719, UGT2B7 rs28365063, UGT2B7 rs7439366, UGT1A9 rs2741049, UGT1A6 rs6759892.

[0017] Preferably, in step (2):

[0018] The blood concentration of the active metabolite MHD of oxcarbazepine is detected by homogeneous enzyme immunoassay.

[0019] The genotyping of the genetic polymorphism sites is performed by the iPLEX Gold Assay method of the MassARRAY platform of Agena Company, and the genotypes are selected.

[0020] The selected genotypes are subjected to Hardy-Weinberg balance (HWE) test, and if the genotype frequency does not comply with the HWE law, i.e., when the chi-square test P<0.05, it is indicated that the population genetic inheritance is unbalanced, and the stepwise method is no longer used to investigate whether it has a significant influence on the pharmacokinetic parameters.

[0021] Preferably, in step (3):

[0022] The absorption rate constant of the one-compartment model of the primary absorption and elimination is 0.83 / h.

[0023] Preferably, in step (4): the scatter plot and box plot of the individual clearance rate estimated by the Bayesian method plotted against the included covariates are helpful for identifying the covariates with significant influence.

[0024] The covariates include continuous covariates and categorical covariates; the continuous covariates include age, body weight, albumin, glutamic-pyruvic transaminase, hemoglobin, hematocrit, creatinine, oxcarbazepine daily dose; the categorical covariates include gender, concomitant medication, genetic polymorphism; the concomitant medication includes valproic acid, levetiracetam, clonazepam; the genetic polymorphism includes ABCB1 rs1128503, ABCB1 rs3789243, ABCB1 rs1045642, ABCC2 rs2273697, ABCC2 rs717620, ABCC2 rs3740066, ADORA2A rs2298383, AS3MT rs7085104, IL1B rs16944, MTHFR rs1801131, MTHFR rs1801133, SCN1A rs6730344, SCN1A rs2298771, SCN1A rs10167228, SCN1A rs3812718, SCN1A rs6732655, SCN2A rs2304016, SCN2A rs17183814, UGT2B7 rs7668258, UGT2B7 rs7668282, UGT2B7 rs12233719, UGT2B7 rs28365063, UGT2B7 rs7439366, UGT1A9 rs2741049, UGT1A6 rs6759892.

[0025] The final model, in particular:

[0026]

[0027]

[0028] V / F(L) = 17.8;

[0029] wherein CL / F is the clearance, V / F is the apparent volume of distribution, Age is the age (years), BW is the body weight (kg), ABCC2 is the ABCC2 rs2273697 polymorphism, ABCC2 GG = 1, ABCC2 AG = 0 is the wild type patient, ABCC2 AG = 1, ABCC2 GG = 0 is the heterozygous AG genotype patient, ABCC2 GG = 0, ABCC2 AG = 0 is the homozygous AA genotype patient.

[0030] Preferably, in step (5):

[0031] The graphical method includes goodness-of-fit plot and visual predictive check;

[0032] The statistical method is bootstrapping.

[0033] The goodness of fit plot (GOF plot): the goodness of fit plots of the base model and the final model are drawn respectively, mainly including scatter plots of observed concentration value-population predicted value, observed concentration value-individual predicted value, conditional weight residual-time and conditional weight residual-population predicted value. The GOF plot can intuitively evaluate whether the predicted value can well describe the central tendency and dispersion degree of the data, and evaluate the fitting degree of the model through the correlation between the predicted value and the measured value, the distribution range of the residual error, and whether the residual error is uniformly distributed on both sides of the x-axis.

[0034] The visual predictive check (VPC plot): a simulated data set is generated according to the parameter estimate value of the final model, the percentiles (5%, 50% and 95%) of the observed data set and the simulated data set at each time point are calculated respectively, and the closeness and distribution characteristics of the two are compared. If the observed value falls within the 95% confidence interval (CI) range of the simulated data, it indicates that the model prediction performance is good. The VPC plot can intuitively compare the degree of coincidence between the model fitting value and the measured value, and serve as a reference for auxiliary evaluation of the accuracy and prediction ability of the model.

[0035] The bootstrap method: 1000 data sets are generated by taking samples with replacement from the original modeling data, with 1000 sampling times; the final model is fitted based on the 1000 data sets, the number of successful fittings is counted, the robustness of the model is calculated, the median value and 95% CI of each parameter estimate are summarized and compared with the fitting results of the final model, and then the reliability and stability of the parameter estimate value of the model are evaluated. The higher the robustness of the model, and the more the parameters of the final model fall within the 95% CI of the corresponding parameters of the bootstrap method, the better the internal stability and prediction reliability of the model.

[0036] Preferably, in step (6), the number of virtual patients is ≥1000.

[0037] The drug is oxcarbazepine.

[0038] An oxcarbazepine quantitative pharmacological model based on genetic polymorphism is obtained by the above construction method.

[0039] The oxcarbazepine quantitative pharmacological model based on genetic polymorphism is specifically:

[0040]

[0041]

[0042] V / F(L) = 17.8;

[0043] Wherein, CL / F is the clearance rate, V / F is the apparent distribution volume, Age is the age (years), BW is the body weight (kg), ABCC2 is the ABCC2 rs2273697 polymorphism, ABCC2 GG=1, ABCC2 AG=0 is the wild type patient, ABCC2 AG=1, ABCC2 GG=0 is the heterozygous AG genotype patient, and ABCC2 GG=0, ABCC2 AG=0 is the homozygous AA genotype patient.

[0044] The estimated parameter values of the oxcarbazepine quantitative pharmacology model based on genetic polymorphism are shown in Table 1:

[0045] Table 1 Estimated parameter table of the final model

[0046]

[0047]

[0048] Note: RSE is the relative standard error.

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

[0050] The present application provides a construction method of an oxcarbazepine quantitative pharmacology model based on genetic polymorphism. The oxcarbazepine quantitative pharmacology model based on genetic polymorphism constructed by the method has stable and accurate prediction performance. The body weight and the genetic polymorphism ABCC2 rs2273697 are determined as effective covariates affecting the MHD CL / F, so that the inter-individual variation in the MHD pharmacokinetic parameters is better understood. Further, an oxcarbazepine dose optimization scheme is established based on the patient's body weight and genotype, so as to improve the therapeutic effect and avoid the occurrence of adverse reactions. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 It is a technical roadmap for the population pharmacokinetic analysis of oxcarbazepine in Example 1 of the present application.

[0052] Figure 2 It is a box plot of the genetic polymorphism ABCC2 rs2273697 different genotypes and the MHD individual clearance rate estimated by the Bayesian method of the final model in Example 1 of the present application.

[0053] Figure 3 It is a goodness-of-fit plot of the base model and the final model in Example 1 of the present application, wherein (A) is the observed concentration vs. population predicted value; (B) is the observed concentration vs. individual predicted value; (C) is the conditional weighted residual vs. population predicted value; (D) is the conditional weighted residual vs. time.

[0054] Figure 4 For the visualization of the final model in Example 1 of the present application, the black dots represent the observed concentrations; the black lines from top to bottom represent the 5th, 50th, 95th percentiles of the observed values, and the shaded area represents the 95% confidence interval for each line based on simulations.

[0055] Figure 5 For the oxcarbazepine dose simulation results plot of the final model in Example 1 of the present application. DETAILED DESCRIPTION

[0056] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0057] In addition, if the present application involves the description of "first", "second", etc. in the embodiments, the description of "first", "second", etc. is only for the purpose of description, and cannot be understood as indicating or implying the relative importance of the indicated technical features or implicitly indicating the number of the indicated technical features. Therefore, the features limited by "first", "second" can explicitly or implicitly include at least one of the features. In addition, the technical solutions of each embodiment can be combined with each other, but it must be based on the fact that the technical solutions can be realized by those skilled in the art. When the combination of technical solutions appears contradictory or unachievable, it should be considered that the combination of technical solutions does not exist, and is not within the scope of protection required by the present application.

[0058] At present, PPK uses mathematical modeling and simulation technology to conduct population analysis on the scattered blood concentration data in clinical routine monitoring, integrates relevant information such as patients, drugs and diseases, can quantitatively analyze the influence of individual differences (such as factors such as gender, age, weight, combined medication, related genotypes, etc.) on the pharmacokinetics of oxcarbazepine, and combined with the individual characteristics of the patient and the treatment target, to develop the best individualized dosing regimen.

[0059] The polymorphism of drug metabolism enzymes and other related genes (including UGTs, ABC transporters, etc.) is an important factor for the individual differences in the pharmacokinetics of the active metabolite MHD of oxcarbazepine. Therefore, the present application uses a nonlinear mixed effect model method to establish a PPK model of MHD, quantitatively investigates the influence of genetic polymorphism and other influencing factors on the pharmacokinetic parameters of MHD, and provides an important reference for guiding individualized oxcarbazepine medication.

[0060] At present, children with epilepsy have underdeveloped metabolic enzyme systems and organ functions in the body, and there is a disadvantage that a large individual difference in blood drug concentration after oral administration of the same dose of oxcarbazepine causes treatment failure or serious adverse reactions. In order to solve the above technical problems, the present application proposes a method for constructing an oxcarbazepine quantitative pharmacological model based on genetic polymorphism. The present application adopts a non-linear mixed effect model method, combines classical pharmacokinetic theory and statistical principles to establish a population pharmacokinetic model of the active metabolite MHD of oxcarbazepine, and quantitatively investigates the influence of physiological factors, combined drug use and genetic factors on the pharmacokinetic characteristic parameters. The final model uses the goodness-of-fit plot, bootstrap method and visual predictive check for internal validation and evaluation. In the application of the model, the Monte Carlo simulation method is used to predict the MHD trough concentration under different dosing regimens, and dose recommendations are proposed for different subgroups of patients, so as to realize safe, effective and individualized drug use.

[0061] Although population pharmacokinetic models of oxcarbazepine in pediatric populations have been reported in domestic and foreign research, they mainly involve the relationship between demographic factors and the variability of oxcarbazepine pharmacokinetic parameters, few people pay attention to the influence of genetic polymorphism, and no recommended dose based on genetic factors has been proposed. It is worth noting that the polymorphism of drug metabolism enzymes and other genes is an important factor for the individualization of pharmacokinetics, therefore, the present application adopts a non-linear mixed effect model method to construct a population pharmacokinetic model of oxcarbazepine in children with epilepsy, and quantitatively investigates the influence of various factors including genetic polymorphism on the pharmacokinetic characteristic parameters of oxcarbazepine; and the stability and prediction performance of the model are comprehensively evaluated, and the Monte Carlo simulation method is used to develop a reasonable dosing regimen for children, guiding individualized drug use.

[0062] Embodiment 1

[0063] As Figure 1 shown, a method for constructing an oxcarbazepine quantitative pharmacological model based on genetic polymorphism adopts a population pharmacokinetic model method, and the present embodiment specifically includes the following steps:

[0064] I. Design of research plan

[0065] Population pharmacokinetic analysis is adopted, strict inclusion and exclusion criteria are formulated, and genetic polymorphism sites related to drug transporters, drug metabolism and oxcarbazepine efficacy are selected, and patient data meeting the conditions are included in the analysis; the specific method is as follows:

[0066] Population pharmacokinetic analysis was performed using NONMEM software (Version 7.5, ICON Development Solution, MD, USA). The blood concentration of oxcarbazepine was determined by homogeneous enzyme immunoassay, and the genotyping experiment was performed by Agena MassARRAY platform iPLEX Gold Assay method.

[0067] The inclusion criteria were: 1) Children under 12 years old who met the latest seizure type classification of the International League Against Epilepsy (ILAE); 2) Patients who strictly followed the doctor's advice or took oxcarbazepine according to the requirements of this study; 3) Patients who did not receive special treatment such as dialysis or diuresis during the study period that significantly affected drug elimination. The exclusion criteria were: 1) Patients with missing key study data: such as unknown dose, no recent weight record, abnormal weight change; 2) Patients with unclear blood concentration monitoring information or sampling time not meeting the requirements; 3) Patients with significant abnormalities in liver and kidney function during medication. This study protocol was approved by the Medical Ethics Committee of Shenzhen Bao'an Maternal and Child Health Hospital, and the patient and their guardian had signed the informed consent form.

[0068] II. Data collection

[0069] The basic information of children with epilepsy treated with oxcarbazepine was collected retrospectively: (1) General information: name, registration number, date of birth, etc. (2) Demographic data: gender, age, height and weight, etc.; (3) Biochemical indicators: albumin, alanine aminotransferase, hemoglobin, hematocrit, creatinine, etc.; (4) Drug use: oxcarbazepine dosage form, dosing regimen, oxcarbazepine daily dose, date of starting oxcarbazepine and changing oxcarbazepine dose, dosing time and blood sampling time; (5) Concomitant medication: other antiepileptic drugs such as valproic acid, levetiracetam, clonazepam, etc.; (6) Oxcarbazepine blood concentration detection data; (7) Genotype selected in this study for genetic polymorphism.

[0070] The data was arranged into a comma-separated plain text format ".csv" file according to the format requirements of NONMEM software, and a uniform table (Table 2) was made, which contained two-dimensional data of columns and rows, and the accuracy of the data was checked. Missing values, outliers and abnormal values were processed, and the data was entered and checked by two people to ensure the reliability and accuracy of the data. Then exploratory data analysis was performed: scatter plot of blood concentration and time was drawn to visually understand the distribution of patient blood concentration and preliminarily estimate the pharmacokinetic characteristics of oxcarbazepine; the distribution of data was observed, and the median, minimum, maximum, mean and standard deviation of each variable were calculated.

[0071] Table 2 Example of NONMEM data file information collection for oxcarbazepine

[0072]

[0073] A total of 320 blood samples were collected from 91 children with epilepsy treated with oxcarbazepine. Table 3 shows the main demographic characteristics of the patients and the concomitant medications.

[0074] Table 3 Demographic information of children with epilepsy

[0075]

[0076]

[0077] Hardy-Weinberg equilibrium (HWE) test was performed on the selected genotypes. If the genotype frequency was in line with HWE law, i.e. the chi-square test P>0.05, it indicated that the population genetic balance was in equilibrium, and the data came from the same Mendelian population. For the genotypes not in HWE balance, the stepwise method was not used to investigate whether they had a significant impact on the pharmacokinetic parameters (Table 4).

[0078] Table 4 Distribution of selected genotype frequencies and HWE balance judgment

[0079]

[0080]

[0081] SCN1A rs6732655 did not meet the Hardy-Weinberg equilibrium and was not included in the covariate screening. The remaining 24 SNPs were in line with the Hardy-Weinberg equilibrium.

[0082] Three, base model

[0083] The base model is a model that characterizes the overall characteristics of the data, including a structural model and a random effects model. The structural model is a classic pharmacokinetic model. On the basis of the selected structural model, further analysis is carried out. The inter-individual variation model uses an exponential model to fit. The residual variation model examines the additive, proportional, and mixed types and selects the best one. The estimation method of the model pharmacokinetic parameters uses the first-order conditional estimation method with inter-individual-residual variation interaction (FOCE-I).

[0084] In this study, the samples collected were from routine blood concentration monitoring results, and most patients were blooded before taking the medicine (steady-state trough concentration). Therefore, a one-compartment model with first-order absorption and elimination (ADVAN2TRANS2) was used for modeling. The absorption rate constant (K a ) was set to 0.83 / h. The lack of absorption and distribution phase information would lead to a large inter-individual variation in the estimated V / F. Therefore, V / F was fixed. The inter-individual variation model uses an exponential model to fit, and the residual variation model uses a mixed type to fit.

[0085] Four, fixed effect model

[0086] After the base model was constructed, covariate model was established using forward inclusion method / reverse elimination method.

[0087] The covariates were screened to describe and explain the source of variation of pharmacokinetic parameters; firstly, the individual parameters were calculated by using Bayesian principle method, then the individual parameters were plotted with the covariates to be investigated, the covariates with trend distribution were included in further statistical test; several fixed effect parameters which obviously affected the pharmacokinetics of oxcarbazepine were found from various influencing factors, after the full model was established, reverse elimination method was used to investigate each influencing factor, and the final model was obtained after eliminating the fixed effect parameters without significant meaning.

[0088] Forward inclusion method: each covariate was added to the base model to investigate the change of objective function value (OFV); if the addition of a covariate made the model OFV decrease more than 3.84 (χ 2 , P<0.05), indicating that the factor could significantly improve the goodness of fit of the model, then the covariate was retained, otherwise it was eliminated; after all the covariates were added to the base model, the model with the most decreased OFV value was taken as the base model of the next round; in this way, all the covariates were screened, and finally the full regression model was obtained; the selection of the model should also consider the degree of reduction of the inter-individual variation of the parameters.

[0089] Reverse elimination method: the model after forward inclusion might include some unnecessary covariates, leading to unstable model or over-parameterization; therefore, based on the full regression model, only one covariate was eliminated in each round, and the changes of model parameters and OFV value were observed to investigate the necessity of the factor in the model; after the reverse elimination process, the final model was obtained by retaining all the covariates with OFV increase value more than 6.63 (χ 2 , P<0.01); during the process of obtaining the final model, if the condition number exceeded 1000, it indicated that the model was over-fitted, and the model structure should be simplified or the model parameter value should be fixed to improve the accuracy of the estimation of pharmacokinetic parameters.

[0090] After the base model was constructed, covariate model was established using forward inclusion method / reverse elimination method. The full regression model was established with each continuous covariate (age, body weight, albumin, hemoglobin, creatinine, oxcarbazepine daily dose, etc.) and categorical covariate (gender, concomitant medication and genotype) with statistical difference, and the specific screening process was shown in Table 5. The combination of more than 5% of the proportion was evaluated, since there was a strong correlation between age and body weight, their effects on the clearance rate of oxcarbazepine should be investigated respectively and repeated inclusion in the model should be avoided. Figure 2Box plot of individual clearance calculated based on Bayesian algorithm for ABCC2 rs2273697 different genotypes in this study, showing the distribution trend of the polymorphism on CL / F.

[0091] Table 5 Analysis of the effect of covariates on CL / F of MHD

[0092]

[0093] First round forward inclusion

[0094] 1 Base model 1304.159

[0095] 2 Include weight in model 1 1231.236 -72.923 <0.001

[0096] 3 Include age in model 1 1240.726 -63.433 <0.001

[0097] 4 Include daily dose in model 1 1300.184 -3.975 <0.05

[0098] 5 Include gender in model 1 1304.046 -0.113 NS

[0099] 6 Include total protein in model 1 1297.208 -6.951 <0.01

[0100] 7 Include albumin in model 1 1299.537 -4.622 <0.05

[0101] 8 Include glutamic-pyruvic transaminase in model 1 1304.879 0.720 NS

[0102] 9 Include serum creatinine in model 1 1279.669 -24.490 <0.001

[0103] 10 Include hemoglobin in model 1 1300.188 -3.971 <0.05

[0104] 11 Include hematocrit in model 1 1305.872 1.713 NS

[0105] 12 Include levetiracetam in model 1 1303.347 -0.812 NS

[0106] 13 Include valproic acid in model 1 1303.970 0.189 NS

[0107] 14. Clonazepam on the basis of model 1 1299.275 -4.884 <0.05

[0108] 15. ABCB1 rs1128503 on the basis of model 1 1302.953 -1.206 NS

[0109] 16. ABCB1 rs3789243 on the basis of model 1 1303.646 -0.513 NS

[0110] 17. ABCB1 rs1045642 on the basis of model 1 1303.823 -0.336 NS

[0111] 18. ABCC2 rs2273697 on the basis of model 1 1299.005 -5.514 <0.05

[0112] 19. ABCC2 rs717620 on the basis of model 1 1303.909 -0.250 NS

[0113] 20. ABCC2 rs3740066 on the basis of model 1 1304.134 -0.025 NS

[0114] 21. ADORA2A rs2298383 on the basis of model 1 1300.137 -3.822 NS

[0115] 22. AS3MT rs7085104 on the basis of model 1 1303.683 -0.476 NS

[0116] 23. IL1B rs16944 on the basis of model 1 1303.292 -0.867 NS

[0117] 24. MTHFR rs1801131 on the basis of model 1 1300.678 -3.481 NS

[0118] 25. MTHFR rs1801133 on the basis of model 1 1295.899 -8.260 <0.01

[0119] 26. SCN1A rs6730344 on the basis of model 1 1302.749 -1.410 NS

[0120] 27. Incorporate SCN1A rs2298771 into model 1 1301.086 -3.073 NS

[0121] 28. Incorporate SCN1A rs10167228 into model 1 1300.854 -3.305 NS

[0122] 29. Incorporate SCN1A rs3812718 into model 1 1303.600 -0.559 NS

[0123] 30. Incorporate SCN2A rs2304016 into model 1 1303.183 -0.976 NS

[0124] 31. Incorporate SCN2A rs17183814 into model 1 1302.047 -2.112 NS

[0125] 32. Incorporate UGT2B7 rs7668258 into model 1 1300.402 -3.757 NS

[0126]

[0127] After a series of forward inclusion and backward elimination model fitting and screening of covariates, the weight segmentation model with 2 years of age as the cut point had a better fitting effect on the data, and the weight and gene polymorphism ABCC2 rs2273697 were retained in the final model, and the final model of MHD was obtained:

[0128]

[0129]

[0130] V / F(L) = 17.8.

[0131] Wherein, CL / F is the clearance rate, V / F is the apparent distribution volume, Age is the age (years), BW is the body weight (kg), ABCC2 is the ABCC2 rs2273697 polymorphism, ABCC2 GG = 1, ABCC2 AG = 0 is the wild type patient, ABCC2 AG = 1, ABCC2 GG = 0 is the heterozygous AG genotype patient, ABCC2 GG = 0, ABCC2 AG = 0 is the homozygous AA genotype patient. The condition number is 391, and the model shows no signs of over-parameterization.

[0132] Table 6. Estimates of parameters of the base model, final model and bootstrap verification

[0133]

[0134] a 1000 bootstrap runs, 967 were successful, used to calculate point estimates and 95% CI.

[0135] b Relative error % = (bootstrap median - final model parameter estimate) / final model parameter estimate x 100%.

[0136] V. Model evaluation

[0137] The stability and predictive performance of the model were evaluated using graphical methods (goodness-of-fit plots, visual predictive check) and statistical methods (bootstrap).

[0138] Goodness-of-fit plots (GOF plots): GOF plots were generated for both the base model and final model. The plots included scatter plots of observed concentration vs. population predicted value, observed concentration vs. individual predicted value, conditional weighted residuals vs. time, and conditional weighted residuals vs. population predicted value. The GOF plots provided a visual assessment of whether the predicted values adequately described the central tendency and dispersion of the data, and the degree of fit of the model was assessed by the correlation between the predicted and observed values, and the distribution of the residuals, which should be evenly distributed on both sides of the x-axis.

[0139] Visual predictive check (VPC plot): Simulated data sets were generated based on the parameter estimates of the final model. The percentiles (5%, 50%, and 95%) of the observed data set and simulated data sets were calculated at each time point, and the closeness of the two and the distribution characteristics were compared. If the observed values fell within the 95% confidence interval (CI) of the simulated data, it indicated that the model had good predictive performance. The VPC plot provided a visual comparison of the degree of overlap between the fitted values and the observed values, which served as a reference for evaluating the accuracy and predictive ability of the model.

[0140] Bootstrap: Resampling with replacement was performed on the original modeling data, with 1000 samplings to generate 1000 data sets. The final model was fitted to the 1000 data sets, and the number of successful fittings was counted to calculate the robustness of the model. The median value and 95% CI of each parameter estimate were summarized and compared with the results of the final model fitting, to evaluate the reliability and stability of the parameter estimates. The higher the robustness of the model, and the closer the parameter estimates of the final model to the 95% CI of the corresponding parameters in the bootstrap method, the better the internal stability and predictive reliability of the model.

[0141] The goodness-of-fit plots based on the base model and final model were compared Figure 3), the trend lines of the final model population prediction and individual prediction were highly consistent with the reference line, the goodness of fit was obviously improved, and the central tendency of the data was well described; most of the condition weighted residuals were symmetrically distributed on both sides of the reference line (y=0) and distributed between-2 and +2, without obvious deviation or significant trend; the above indicated that the prediction effect of the model was better after the two covariates of body weight and genotype ABCC2 rs2273697 were included. The results of the visual predictive check of the final model are shown in Figure 4 The 5th, 50th and 95th percentiles of the observation results were approximately distributed in the 95% CI of the simulated concentration of each interval, indicating that the model had good prediction performance. The robust rate of the bootstrap analysis was 96.7%, the parameter estimation value of the final model was included in the 95% CI range of the model parameters obtained by the bootstrap method, and the relative error of the corresponding median was less than 4.4%, which indicated that the accuracy and robustness of the parameter estimation of the final model (Table 6).

[0142] Six, dose optimization

[0143] It is a necessary condition to ensure the therapeutic safety and effectiveness of oxcarbazepine to select a suitable administration scheme to maintain the steady-state blood drug concentration in the therapeutic window. In order to make the final model more directly serve the clinic, the simulation module in NONMEM was used to perform Monte Carlo simulation on patients with different body weights and ABCC2 rs2273697 genetic polymorphisms, and the steady-state trough concentration simulation results of 1,000 virtual patients under the administration scheme of oxcarbazepine 10-70 mg / kg / day were observed Figure 5 , which has important reference value for administration scheme optimization.

[0144] According to the included covariates, the virtual patients (n=1,000) were divided into different subgroups, and based on the final model, the steady-state trough concentration of MHD under different dose schemes was calculated by using the Monte Carlo simulation method.

[0145] The present application is based on the oxcarbazepine quantitative pharmacology model of genetic polymorphism, that is, the MHD pediatric population pharmacokinetic model, and specifically comprises:

[0146]

[0147]

[0148] V / F(L)=17.8.

[0149] Wherein, CL / F is the clearance rate, V / F is the apparent distribution volume, Age is the age (years), BW is the body weight (kg), ABCC2 is the ABCC2 rs2273697 polymorphism, ABCC2 GG=1, AG=0 is the wild type patient, ABCC2 AG=1, GG=0 is the heterozygous AG genotype patient, and ABCC2 GG=0, AG=0 is the homozygous AA genotype patient.

[0150] The final population pharmacokinetic model parameter table is shown in Table 1.

[0151] The results show that the MHD trough concentration gradually increases with the increase of the dose, and the children over 12 kg (2 years old) are more sensitive to the change of the drug dose, and there is a risk of high exposure when the dose is up to 40-60 mg / kg / day. In the patients with ABCC2 rs2273697 homozygous (AA genotype) and heterozygous (AG genotype) variants, the CL / F is reduced by 25% and 14% respectively compared with the wild homozygous type (GG genotype). Therefore, for patients carrying variant alleles, a lower maintenance dose is needed for the same body weight. Taking a typical patient of 10.5 years old and 35 kg as an example, the probability of reaching the target blood drug concentration range is higher for the GG genotype patient at a dose of 40 mg / kg / day, and the patient carrying the variant allele is recommended to be given a smaller dose of 30 mg / kg / day to prevent the concentration from exceeding the toxicity threshold.

[0152] Compared with the prior art, the present application has the beneficial effects that: the present application provides a construction method of an oxcarbazepine quantitative pharmacology model based on genetic polymorphism, and the oxcarbazepine quantitative pharmacology model based on genetic polymorphism obtained by construction has stable and accurate prediction performance; the body weight and the genetic polymorphism ABCC2 rs2273697 are determined as effective covariates affecting the MHD CL / F, and the inter-individual variation in the MHD pharmacokinetic parameters is better understood; and further based on the patient body weight and the genotype, an oxcarbazepine dose optimization scheme is established to improve the treatment effect and avoid the occurrence of adverse reactions.

[0153] The above merely describes preferred embodiments of the present application, but should not be used to limit the present application, and any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for constructing an oxcarbazepine pharmacokinetic model based on genetic polymorphism, characterized by comprising the steps of: Comprising the following steps: ​ (1) Design research plan: adopt population pharmacokinetic analysis, formulate the inclusion and exclusion criteria, select the polymorphic sites of genes related to drug transporters, drug metabolism and oxcarbazepine efficacy; (2) Collect data: collect the data including demographic information, biochemical indicators, medication, combined medication, blood concentration detection data of oxcarbazepine active metabolite MHD, and genotyping of polymorphic sites of genes; (3) Basic model: use a one-compartment model with first-order absorption and elimination for modeling; The inter-individual variability model is fitted by an exponential model, and the residual variation model is fitted by a mixed type; (4) Fixed effect model: use forward inclusion method / reverse elimination method to establish the covariate model; from various influencing factors, select the fixed effect parameters that significantly affect the pharmacokinetics of oxcarbazepine active metabolite MHD, plot the covariates to be investigated and the individual clearance rate estimated by empirical Bayes method; obtain the final model; body weight and gene polymorphism ABCC2 rs2273697 are retained in the final model; (5) Model evaluation: evaluate the stability and prediction performance of the model by graphical and statistical methods; (6) Dose optimization: according to the included covariates, divide the virtual patients into different subgroups, and based on the final model, use Monte Carlo simulation method to predict the distribution of steady-state blood drug concentration under different dosing regimens, and complete the model construction; In step (4): the final model is specifically: if Age > 2 years old, ； if Age ≤ 2 years old, ； ； Wherein, CL / F is the clearance rate, the unit of clearance rate is L / h, V / F is the apparent distribution volume, the unit of apparent distribution volume is L, Age is age, BW is body weight, ABCC2 is ABCC2 rs2273697 polymorphism, ABCC2 GG=1, ABCC2AG=0 is wild type patient, ABCC2 AG=1, ABCC2GG=0 is heterozygous AG genotype patient, ABCC2 GG=0, ABCC2AG=0 is homozygous AA genotype patient, and η is the inter-individual variation of clearance rate.

2. The method for constructing an oxcarbazepine pharmacokinetic model based on genetic polymorphism according to claim 1, characterized by: In step (1): the inclusion and exclusion criteria include the inclusion criteria and the exclusion criteria; the inclusion criteria are: 1) children less than 12 years old with seizure types in accordance with the International League Against Epilepsy; 2) patients who strictly follow the doctor's advice according to the diagnosis; 3) patients who do not receive dialysis or special treatment that significantly affects drug elimination during the study; the exclusion criteria are: 1) patients with missing key research data, including unknown dose, no recent weight record, and abnormal weight change; 2) patients with unclear blood drug concentration monitoring information or sampling time not meeting the requirements; 3) patients with significant abnormalities in liver and kidney function during medication.

3. The method for constructing an oxcarbazepine pharmacokinetic model based on genetic polymorphism according to claim 1, characterized by: In step (1): the genetic polymorphism sites include ABCB1 rs1128503, ABCB1 rs3789243, ABCB1 rs1045642, ABCC2 rs2273697, ABCC2 rs717620, ABCC2 rs3740066, ADORA2Ars2298383, AS3MT rs7085104, IL1B rs16944, MTHFR rs1801131, MTHFR rs1801133, SCN1Ars6730344, SCN1A rs2298771, SCN1A rs10167228, SCN1A rs3812718, SCN1A rs6732655, SCN2A rs2304016, SCN2A rs17183814, UGT2B7 rs7668258, UGT2B7 rs7668282, UGT2B7rs12233719, UGT2B7 rs28365063, UGT2B7 rs7439366, UGT1A9 rs2741049, UGT1A6rs6759892.

4. The method for constructing an oxcarbazepine pharmacokinetic model based on genetic polymorphism according to claim 1, characterized by: In step (2): the blood concentration of the oxcarbazepine active metabolite MHD is detected by homogeneous enzyme immunoassay; The genotyping of the genetic polymorphism sites is performed by the Agena MassARRAY platform iPLEX Gold Assay method, and genotypes are selected; The selected genotypes are subjected to Hardy-Weinberg balance test, if the genotype frequency does not comply with the HWE law, chi-square test is performed, when the chi-square test P<0.05, it is indicated that the population genetic imbalance, and the stepwise method is no longer used to investigate whether it has a significant influence on the pharmacokinetic parameters.

5. The method for constructing an oxcarbazepine pharmacokinetic model based on genetic polymorphism according to claim 1, characterized by: In step (3): the absorption rate constant of the one-compartment model of the first absorption and elimination is 0.83 / h.

6. The method for constructing an oxcarbazepine pharmacokinetic model based on genetic polymorphism according to claim 1, characterized by: In step (4): the individual clearance rate estimated by the Bayesian method is plotted into a scatter plot and a box plot against the included covariates; The covariates include continuous covariates and categorical covariates; the continuous covariates include age, weight, albumin, alanine aminotransferase, hemoglobin, hematocrit, creatinine, oxcarbazepine daily dose; the categorical covariates include gender, concomitant medication, genetic polymorphism; the concomitant medication includes valproic acid, levetiracetam, clonazepam; the genetic polymorphism includes ABCB1 rs1128503, ABCB1 rs3789243, ABCB1 rs1045642, ABCC2 rs2273697, ABCC2 rs717620, ABCC2 rs3740066, ADORA2A rs2298383, AS3MT rs7085104, IL1B rs16944, MTHFR rs1801131, MTHFR rs1801133, SCN1A rs6730344, SCN1A rs2298771, SCN1A rs10167228, SCN1A rs3812718, SCN1A rs6732655, SCN2A rs2304016, SCN2A rs17183814, UGT2B7 rs7668258, UGT2B7 rs7668282, UGT2B7 rs12233719, UGT2B7 rs28365063, UGT2B7 rs7439366, UGT1A9 rs2741049, UGT1A6 rs6759892.

7. The method for constructing an oxcarbazepine pharmacokinetic model based on genetic polymorphism according to claim 1, characterized by: In step (5), the graphic method includes goodness-of-fit plot and visual predictive check; the statistical method is bootstrapping method; In step (6), the number of virtual patients is greater than or equal to 1000. The drug is oxcarbazepine.

Images