Methods and systems for determining population pharmacokinetic models of propofol and its derivatives

Abstract

Provided are a method and system for determining a population pharmacokinetic model of propofol and derivatives thereof. The method comprises determining a formula as a population pharmacokinetic model of a compound of formula (I) or propofol: wherein the formula of the pharmacokinetic parameters in the population pharmacokinetic model of the compound of formula (I) comprises: and the formula of the pharmacokinetic parameters in the population pharmacokinetic model of propofol comprises: formula (I).

Description

Technical Field

[0001] This invention relates to the pharmaceutical field, and more specifically, to a method and system for determining population pharmacokinetic models of propofol and its derivatives. Background Technology

[0002] Propofol derivative emulsion injection (hereinafter referred to as propofol derivative, chemical name 2-[(1R)-1-cyclopropylethyl]-6-isopropylphenol) is a novel intravenous anesthetic drug with independent intellectual property rights developed by Sichuan Haisco Pharmaceutical Co., Ltd. It is intended for sedation / anesthesia during various diagnostic examinations or treatments, induction and maintenance of general anesthesia, and sedation of intensive care subjects under mechanical ventilation (ICU sedation). Its active ingredient, propofol derivative, is a chemical entity similar to propofol, a single diastereomer with two R-type chiral centers. The pharmaceutical design strategy is to systematically improve the pharmacological and physicochemical properties of the drug's binding to receptors, resulting in a propofol derivative compound superior to propofol. The main mechanism of action of propofol derivative is through enhancing the ion channel mediated by γ-aminobutyric acid type A (GABAA) receptors, causing chloride ion influx, thereby achieving central nervous system inhibition. This channel is also the main target of propofol. Propofol derivatives exhibit rapid onset of action and stable, rapid recovery. Furthermore, they demonstrate higher selectivity and in vitro / in vivo activity, with a potency 4-5 times that of propofol. Animal studies have shown more stable hemodynamics. Additionally, under the same conditions and concentrations, ultrafiltration revealed lower free drug concentrations of propofol derivatives compared to propofol (Jing An®) in the aqueous phase, suggesting a potential reduction or elimination of injection site pain. Summary of the Invention

[0003] The purpose of this invention is to provide a method for determining a population pharmacokinetic model of propofol derivative drugs.

[0004] The purpose of this invention is to provide a method that can quantitatively assess the impact of intrinsic and extrinsic factors on drug-pharmaceutical (PK) ratios. The population pharmacokinetic (PopPK) model established according to this invention can estimate individual exposure levels for use in exposure-response (ER) studies. ER analysis is crucial for understanding drug safety and efficacy. Although dose is a commonly used direct indicator of drug exposure in clinical trials, serum / plasma drug concentrations are a more direct indicator of exposure to the drug's target, and thus correlate with clinical efficacy and safety.

[0005] Another object of the present invention is to provide a system for clinical individual dosing parameters of a compound of formula (I) or propofol.

[0006] To achieve the above objectives, in one respect, the present invention provides a method for determining a pharmacokinetic model of a compound (propofol derivative) represented by formula (I), wherein the method includes determining the following formula as the pharmacokinetic model of the compound represented by formula (I): Formula (I) Among them, the pharmacokinetic parameter formulas in the compound population pharmacokinetic model of formula (1) include: ; Among them, CL i SITE represents the central compartment clearance rate of the i-th subject; SITE=0 when the blood is drawn from a vein and SITE=1 when the blood is drawn from an artery; WT represents body weight; TP represents total protein; η CL,i Let η be the inter-individual variation of CL for the i-th subject, and let η follow a normal distribution with a mean of 0 and a variance of ω2, where ω2 is the element on the diagonal of the variance-covariance matrix (Ω) of the inter-individual variation.

[0007] This invention provides a method for determining a population pharmacokinetic model of propofol, wherein the method includes determining the following formula as the population pharmacokinetic model of propofol: The formulas for pharmacokinetic parameters in the propofol population pharmacokinetic model include:

[0008] Among them, CL i η represents the central ventricular clearance rate of the i-th subject. CL,i Let η be the inter-individual variation of CL for the i-th subject, and let η follow a normal distribution with a mean of 0 and a variance of ω2, where ω2 is the element on the diagonal of the variance-covariance matrix (Ω) of the inter-individual variation.

[0009] The above model demonstrates the relationship between drug clearance and relevant covariates (body weight, total protein, and administration site). Clinically, this model can be used to assess the impact of covariates on pharmacokinetic (PK) parameters (primarily drug exposure) to guide clinical dosing. First, based on the PopPK model established in this study, individual PK parameters of the subjects were estimated using a Bayesian post-hoc method. The pharmacokinetic curves of intravenous infusion of propofol derivatives or propofol were simulated based on the actual administered dose, thereby calculating the area under the curve (exposure) at different time points. Simultaneously, correlation data analysis between exposure and efficacy and safety is needed to provide a reasonable dosing regimen.

[0010] The above-mentioned population pharmacokinetic models (PopPK models) of this invention all ultimately selected a three-compartment model with zero-order absorption and first-order linear elimination in the central compartment (e.g., Figure 1As shown), the PopPK model consists of the following parameters: central chamber clearance rate (CL), central chamber distribution volume (V1), peripheral chamber distribution volume (V2, V3), and inter-chamber clearance rate (Q2, Q3).

[0011] According to some specific embodiments of the present invention, the pharmacokinetic parameter formula in the compound population pharmacokinetic model of formula (1) further includes:

[0012] Among them, V 1i This represents the central ventricular distribution volume of the i-th subject. V 2i This represents the peripheral 1 distribution volume of the i-th subject. V 3i This represents the peripheral 2-distribution volume of the i-th subject. Q 2i This represents the interventricular clearance rate between the peripheral chamber 1 and the central chamber for the i-th subject. Q 3i Let represent the peripheral ventricular clearance rate and the central ventricular clearance rate of the i-th subject, respectively. AGE represents age; η represents the inter-individual variation of the corresponding parameter.

[0013] According to some specific embodiments of the present invention, the formulas for pharmacokinetic parameters in the propofol population pharmacokinetic model further include:

[0014] According to some specific embodiments of the present invention, the method for determining the population pharmacokinetic model of the compound shown in formula (I) and propofol includes obtaining the population pharmacokinetic model by utilizing the influence of covariates on the pharmacokinetic parameters in the population pharmacokinetic model of the compound and propofol in formula (I).

[0015] According to some specific embodiments of the present invention, the method for determining the pharmacokinetic models of the compound shown in formula (I) and propofol includes the following steps (the pharmacokinetic models of the compound shown in formula (I) and propofol can both be determined by a method including the following steps): (1) Data acquisition; (2) Determine the data to be included in the analysis; (3) Data processing; (4) Establishment of the initial basic model of population pharmacokinetics; (5) Establishment of the final basic model of population pharmacokinetics; (6) Establishment of a population pharmacokinetic model; (7) Evaluation of population pharmacokinetic models.

[0016] According to some specific embodiments of the present invention, the above steps can be performed sequentially in the order described above.

[0017] According to some specific embodiments of the present invention, the data source for step (1) is clinical trial data.

[0018] According to some specific embodiments of the present invention, step (2) includes determining the pharmacokinetic dataset to be included in the analysis by evaluating the included clinical trial data.

[0019] According to some specific embodiments of the present invention, the data included in the analysis in step (2) include blood drug concentration data, baseline demographic data, blood biochemical index data and blood collection sites.

[0020] According to some specific embodiments of the present invention, the baseline demographic data includes any two or more combinations of race, age, height, weight, and sex; the blood biochemical data includes any two or more combinations of total blood protein content, creatinine clearance, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and total bilirubin.

[0021] According to some specific embodiments of the present invention, step (2) includes determining the pharmacokinetic dataset to be included in the analysis by evaluating the included clinical trial data.

[0022] According to some specific embodiments of the present invention, step (3) includes the identification and processing of combined data of one or more of the following: observations below the detection limit, outlier data, outliers, and missing covariates.

[0023] According to some specific embodiments of the present invention, wherein, The determination and processing of observations below the detection limit include: the detection limit is determined by the detection instrument, observations below the detection limit are not used for population pharmacokinetic analysis, and when the proportion of observations below the detection limit is greater than 15%, the likelihood function method is used to examine the impact of observations below the detection limit on model fitting and modeling parameters. The identification and handling of outlier data includes: checking for outliers in the sample by examining the subject's dosing time and the corresponding plasma concentration curve, and excluding outliers; The identification and handling of outliers includes: identifying outliers based on residual analysis of the initial modeling results, and excluding outliers; Handling of missing covariates: If the missing rate of covariates for subjects is <15%, for continuous covariates, the median in the dataset is used to fill in the missing values; for categorical covariates, the most common category is used to fill in the missing values. If the missing rate of covariates for subjects is >15%, no values ​​are filled in, and subjects with complete covariate information are given Bayesian estimation to conduct exploratory analysis of the PK parameters.

[0024] According to some specific embodiments of the present invention, the data with values ​​below LLOQ in the pharmacokinetic samples account for only 6.8% (167 / 2463), so the M3 method is not required.

[0025] According to some specific embodiments of the present invention, the outlier data includes: 1) There are repeated concentration records at the same time point, except for arterial and venous concentrations at the same time point; 2) The valley concentration is greater than the corresponding peak concentration; 3) The administration time occurs after the peak concentration; 4) The administration time occurs before the trough concentration; 5) Concentration records after intravenous administration, prior to peak concentration; 6) Unexplained sudden drops or rises in concentration.

[0026] According to some specific embodiments of the present invention, the method for confirming detailed information on outliers is as described in the standard PopPK guidelines (the U.S. Food and Drug Administration (FDA) guidelines on industry population pharmacokinetics and the Committee on Human Products for Use (CHMP) guidelines on reporting population pharmacokinetics analysis results).

[0027] According to some specific embodiments of the present invention, outliers are data points outside the normal range of the dataset, which are determined based on residual analysis of the initial modeling results.

[0028] According to some specific embodiments of the present invention, if the absolute value of the conditionally weighted residual (CWRES) exceeds 5, the data point is considered an outlier and removed from the PopPK model. After the final population pharmacokinetic model is determined, outliers are added to the analysis data for model reconstruction to assess whether outliers have any impact on the model.

[0029] According to some specific embodiments of the present invention, step (4) includes comparing multiple structural models based on blood drug concentration and time curves, selecting the best one as the initial structural model, and forming an initial basic model with the residual model.

[0030] According to some specific embodiments of the present invention, the initial structural model is a three-compartment model with zero-order absorption and first-order linear elimination in the central compartment (e.g., Figure 1As shown), the parameters of the initial structural model include: Central chamber clearance rate CL, central chamber distribution volume V1, peripheral chamber 1 distribution volume V2, peripheral chamber 2 distribution volume V3, clearance rate between peripheral chamber 1 and central chamber Q2, clearance rate between peripheral chamber 2 and central chamber Q3, infusion rate R0, and elimination rate constant K.

[0031] According to some specific embodiments of the present invention, the initial basic model includes describing the inter-individual differences in the PK parameter using the following formula:

[0032] in θi This represents the PK parameter of the i-th subject; θ T The natural logarithm of the typical values ​​of the PK parameter in the population; η i The variation among individuals is expressed as follows: the mean is 0 and the variance is ω. 2 A normally distributed random variable, where ω 2 The value represents the element on the diagonal of the variance-covariance matrix of the variation among individuals.

[0033] According to some specific embodiments of the present invention, the initial basic model includes using the following formula to describe the variability of the residuals:

[0034] in y ij This represents the j-th observed concentration of the i-th subject. ij ε represents the model-predicted concentration for the i-th subject. ij Let represent the proportional residual of the j-th observed concentration for the i-th subject. The observed concentration and the predicted concentration are independent of each other and follow normal distributions with mean 0 and variance σ², respectively.

[0035] According to some specific embodiments of the present invention, step (4) includes comparing any two or more of the following structural models based on blood drug concentration and time curves, and selecting the best initial structural model: a one-compartment model, a two-compartment model, and a three-compartment model.

[0036] According to some specific embodiments of the present invention, step (4) includes comparing multiple structural models by comparing whether the objective function decreases significantly between two nested models, and selecting the best initial structural model.

[0037] According to some specific embodiments of the present invention, step (5) includes determining the covariates to be included in the evaluation based on clinical knowledge and drug action mechanism, and establishing a final basic model of population pharmacokinetics based on the covariates to be included in the evaluation.

[0038] According to some specific embodiments of the present invention, the covariates included in the assessment include baseline demographic characteristics covariates, blood biochemical indicators covariates, and blood collection sites.

[0039] According to some specific embodiments of the present invention, the baseline demographic covariates include any two or more combinations of baseline age, sex, weight, and ethnicity; the blood biochemical covariates include any two or more combinations of creatinine clearance, total protein, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and total bilirubin.

[0040] According to some specific embodiments of the present invention, step (6) includes: a) Pre-screening of covariates; b) Use forward and backward methods to finally screen the covariates and establish a population pharmacokinetic model.

[0041] According to some specific embodiments of the present invention, the covariate pre-screening includes: The correlation between PK parameters and various covariates was analyzed using graphical methods. Linear regression was used for continuous covariates, and analysis of variance was used for categorical covariates. Based on the model evaluation dataset, Bayesian methods were used to estimate the parameters of the subjects in the final baseline model and to estimate the impact of covariates on PK parameters.

[0042] According to some specific embodiments of the present invention, the pre-screening of covariates includes, The following formula is used to analyze the correlation between continuous covariates and PK parameters: ; The following formula is used to analyze the correlation between categorical covariates and PK parameters:

[0043] in θ i This represents the PK parameter of the i-th subject; θ pop This represents the typical value of the PK parameter in the subject population; Cov i This represents the continuous covariate value for the i-th subject; Cov pop This represents the median of a continuous variable in the test group; X iLet represent the categorical variable index for the i-th subject, where a value of 0 indicates the category of the covariate with the most common category, and other integer values ​​indicate other categories; k cov This represents a coefficient that describes the magnitude of the influence of the covariate.

[0044] According to some specific embodiments of the present invention, the final screening of covariates includes: Based on the final basic model, a full model is built using the forward method, and then a population pharmacokinetic model is built using the backward method based on the full model. The forward method includes: adding each covariate sequentially to the initial structural model of the final base model in step (5). Based on the log-likelihood ratio test, when adding one covariate, if the objective function value decreases by more than 6.63 based on the criterion of p < 0.01, the newly added covariate is considered significant. Based on the initial structural model, the covariate with the most significant impact is added first (the more the objective function value decreases, the more significant the impact; the most significant impact is the covariate with the largest decrease in objective function value), forming an improved model. Then, the statistically significant covariates selected in the previous step are tested on the improved model. This process is repeated until no significant covariates can be found. The backtracking method includes: a process of deleting covariates one by one based on the full model. If the objective function value increases by more than 10.83 after deleting one covariate, then the deleted covariate is considered significant on the p<0.001 criterion.

[0045] The forward approach: To build a complete model, a stepwise forward addition method is used, adding each covariate sequentially to the base model. Based on the structural model, the most significant covariates are added first to form an improved model. Then, the statistically significant covariates identified in the previous step are tested on the improved model. This process is repeated until no significant covariates are found. In the forward approach, if highly correlated covariates exist, clinically significant covariates are prioritized for examination, while highly correlated covariates are examined last.

[0046] Backward method: This method is the process of gradually removing covariates based on the full model.

[0047] According to some specific embodiments of the present invention, step (7) includes evaluating the population pharmacokinetic model using one or more combinations of the following methods: model fit diagnostic graph (GOF), visual prediction test (pcVPC), bootstrap, and shrinkage.

[0048] Bootstrap is a resampling technique used to evaluate the stability of a model; it is a nonparametric method.

[0049] According to some specific embodiments of the present invention, wherein, The model fitting diagnostic plot includes one or more combinations of the following graphs: population predicted concentration (PRED) versus observed concentration (DV), individual predicted concentration (IPRED) versus observed concentration, conditionally weighted residual (CWRES) versus population predicted concentration, and conditionally weighted residual versus time after the first dose. Visualized prediction testing (visualized prediction testing with predicted value correction) involves simulating 1000 trials using the final model parameters, covariates, and actual doses, and then plotting the predicted results against the measured values ​​to assess whether the population pharmacokinetic model can adequately describe the pharmacokinetic curve of propofol derivatives. The bootstrapping method involves repeatedly fitting the population pharmacokinetic model to 1000 datasets for bootstrapping replication, and randomly selecting subject data and covariates (including concentration, time points, dosing history, and all covariates) to replace them for replication (the number of subjects in each dataset is the same as the original dataset). The contraction method involves estimating individual parameter values ​​of subjects using Bayesian estimation based on a population pharmacokinetic model, and calculating inter-individual variability and individual residuals from model predictions and observed values.

[0050] Visualized prediction tests can compare the consistency between observed values ​​and the median and distribution range (2.5th to 97.5th quantile) of the drug-time curves predicted by the model.

[0051] The contraction method is used to evaluate the inter-individual variability (Ω) and residuals (ε) of the final model, quantifying the estimates of individual parameter values ​​and random errors. If the contraction values ​​of ω and ε are large, such as >30%, then the Bayesian estimation method should be considered with caution.

[0052] According to some specific embodiments of the present invention, the shrinkage method includes using the following formula to evaluate the inter-individual variability and residuals of pharmacokinetic parameters, and quantifying the individual parameter values ​​and random error estimates:

[0053] in, η shrinkage For inter-individual variation, ε shrinkage Let ω represent the individual residuals, and ω represent the inter-individual variation of the individual parameter values ​​estimated by the population pharmacokinetic model. η ph This parameter is for all individuals. η The values ​​are IWRES, which represent the individual weighted residuals, and SD, which represents the standard deviation.

[0054] According to some specific embodiments of the present invention, step (7) further includes estimating the individual PK parameters of the subject using the Bayesian post-hoc method, simulating the drug-time curve of intravenous infusion based on the actual administered dose, and calculating the area under the drug-time curve (AUC0-1 min) from 0 to 1 minute, the area under the drug-time curve (AUC0-2 min) from 0 to 2 minutes, the area under the drug-time curve (AUC0-4 min) from 0 to 4 minutes, the area under the drug-time curve (AUC0-10 min) from 0 to 10 minutes, the area under the drug-time curve (AUC0-24 h) from 0 to 24 hours, and the peak concentration (Cmax).

[0055] According to some specific embodiments of the present invention, the clinical trials included in the analysis of the present invention include: Propofol derivative SAD: A placebo- and positive-drug-controlled, single-dose intravenous escalation trial in healthy Australian subjects; Propofol derivative SAD_02: A phase I escalation trial with a positive drug control and a single intravenous injection in healthy subjects in Australia; Propofol derivative SAD_03: Safety and tolerability trial of a single intravenous injection followed by 30 minutes of continuous intravenous infusion in Australian subjects; Propofol Derivative-101: A single-center, open-label, uncontrolled phase I escalation trial evaluating a single intravenous injection of propofol derivative emulsion in healthy Chinese subjects; Propofol Derivative-103: A single-center, open-label, randomized, two-stage, crossover study evaluating the interaction (DDI) between propofol derivative emulsion injection and rifampicin capsules in healthy Chinese subjects; Propofol Derivative-202: A multicenter, open-label, non-randomized, positive-controlled, dose-escalation phase IIa clinical study evaluating the tolerability, efficacy, and safety of propofol derivative emulsion injection for induction of general anesthesia in subjects undergoing elective surgery. Propofol Derivative-302: This is a phase III, multicenter, randomized, double-blind, parallel-controlled study of propofol to evaluate the efficacy and safety of propofol derivatives compared to propofol in the induction of general anesthesia in Chinese patients undergoing elective surgery.

[0056] According to some specific embodiments of the present invention, a subject may be included in a population pharmacokinetic analysis if at least one well-recorded time of administration and corresponding plasma concentration are collected after administration.

[0057] According to some specific embodiments of the present invention, the method of the present invention is a method for determining a population pharmacokinetic model of propofol derivative emulsion injection.

[0058] According to some specific embodiments of the present invention, the propofol derivative emulsion injection comprises: soybean oil, glycerin, triglycerides, egg yolk lecithin, sodium oleate and sodium hydroxide (see WO / 2016 / 034079).

[0059] According to some specific embodiments of the present invention, the PopPK analysis method is based on the industry guidelines on population pharmacokinetics issued by the U.S. Food and Drug Administration (FDA) and the guidelines on reporting population pharmacokinetic analysis results issued by the Committee on Human Products for Use (CHMP).

[0060] According to some specific embodiments of the present invention, the PopPK analysis method employs a nonlinear mixed-effects model (NONMEM) to estimate the typical values ​​of the parameters and their variability.

[0061] According to some specific embodiments of the present invention, the PopPK analysis software used in the present invention is NONMEM7, version 7.4.0 (ICON Development Solutions, Ellicott City, Maryland, USA); Perl SpeaksNONMEM (PsN), version 3.2.12 (Uppsala University, Sweden).

[0062] The structural models of this invention all ultimately selected a three-compartment model with zero-order absorption and first-order linear elimination in the central compartment, such as... Figure 1 As shown, the PopPK model consists of the following parameters: central chamber clearance rate (CL), central chamber distribution volume (V1), peripheral chamber distribution volume (V2, V3), and inter-chamber clearance rate (Q2, Q3).

[0063]

[0064] Parameters were estimated using the FOCEI method of NONMEM. The correlations between random effects in the underlying PK model were estimated using a diagonal Ω matrix. Based on model diagnostics, a proportional model was selected as the residual model.

[0065] According to some specific embodiments of the present invention, the pre-screening results of step (5) show: The following covariates all had a significant impact on the PK parameter (p<0.01) and were included in the covariate model selection process, where the forward-backward method was used to examine the covariates: (1) Central compartment clearance CL: Creatinine clearance (CLCR), total protein (TP), total bilirubin (TBIL), body weight (WT), age (AGE), race (RACE), and blood collection site (SITE); (2) Central chamber distribution volume V1: age, race and blood collection site.

[0066] According to some specific embodiments of the present invention, in the PopPK model in NONMEM, the stepwise forward addition method (forward method) was used. The results showed that WT, TP, and SITE had a significant impact on CL (p<0.01), and AGE had a significant impact on V1 (p<0.01). Based on this, RACE was examined, and no significant impact on CL or V1 was found. Using the stepwise backward elimination method (backward method), no influencing factors were eliminated (Δ-2LL>10.83).

[0067] On the other hand, the present invention also provides a system for determining the individual dosing parameters of the compound of formula (I) or propofol.

[0068] The drug administration parameters may include individual drug administration data; specifically, such as drug dosage; more specifically, the drug dosage may include an individual's single drug dosage, daily drug dosage, drug administration time, drug administration frequency, etc.

[0069] According to some specific embodiments of the present invention, the system for determining the individual dosing parameters of the compound of formula (I) includes a data acquisition device, a data processing device, and a result output device. Formula (I) The data includes baseline demographic data, blood biochemical index data, and blood collection site information data; the data processing device includes a method for obtaining individual dosing parameters of compound (I) using the following formula: ; CL i SITE represents the central compartment clearance rate of the i-th subject; SITE=0 when the blood is drawn from a vein and SITE=1 when the blood is drawn from an artery; WT represents body weight; TP represents total protein; η CL,i Let η be the inter-individual variation of CL for the i-th subject, and let η follow a normal distribution with a mean of 0 and a variance of ω2, where ω2 is the element on the diagonal of the variance-covariance matrix Ω of the inter-individual variation.

[0070] According to some specific embodiments of the present invention, the result output device is used to output individual drug administration parameter results.

[0071] According to some specific embodiments of the present invention, the data processing device includes obtaining individual dosing parameter results of compound (I) using the following formula: ; V 1i This represents the central ventricular distribution volume of the i-th subject. V 2i This represents the peripheral 1 distribution volume of the i-th subject. V 3i This represents the peripheral 2 distribution volume of the i-th subject. Q 2i This represents the interventricular clearance rate between the peripheral chamber 1 and the central chamber for the i-th subject. Q 3i Let represent the peripheral ventricular clearance rate and the central ventricular clearance rate of the i-th subject, respectively. AGE represents age; η represents the inter-individual variation of the corresponding parameter.

[0072] According to some specific embodiments of the present invention, the system for determining individual propofol dosing parameters includes a data acquisition device, a data processing device, and a result output device; the data includes baseline demographic data, blood biochemical index data, and blood sampling site information data; the data processing device includes obtaining the propofol individual dosing parameter results using the following formula:

[0073] Among them, CL i η represents the central ventricular clearance rate of the i-th subject; CL,i Let η be the inter-individual variation of CL for the i-th subject, and let η follow a normal distribution with a mean of 0 and a variance of ω2, where ω2 is the element on the diagonal of the variance-covariance matrix Ω of the inter-individual variation.

[0074] According to some specific embodiments of the present invention, the data processing device includes obtaining propofol individual dosing parameter results using the following formula:

[0075] Among them, V 1i This represents the central ventricular distribution volume of the i-th subject. V 2i This represents the peripheral 1 distribution volume of the i-th subject. V 3iThis represents the peripheral 2 distribution volume of the i-th subject. Q 2i This represents the interventricular clearance rate between the peripheral chamber 1 and the central chamber for the i-th subject. Q 3i Let represent the peripheral ventricular clearance rate and the central ventricular clearance rate of the i-th subject, respectively. AGE represents age; η represents the inter-individual variation of the corresponding parameter.

[0076] According to some specific embodiments of the present invention, the result output device is used to output individual drug administration parameter results.

[0077] Methods not detailed in this application can be performed by referring to conventional methods in the field.

[0078] In summary, this invention provides a method and system for determining population pharmacokinetic models of propofol derivatives and propofol. The method of this invention has the following advantages: The three-compartment linear pharmacokinetic model effectively described the pharmacokinetic characteristics within the dosage range of this study. Model evaluation showed that the final model possessed good stability, accuracy, and predictive performance; no significant bias was observed in the diagnostic plots; and the pcVPC results demonstrated that the model adequately reproduced the central tendency and variability of the original data. The population pharmacokinetic model of propofol derivatives and propofol established by the method of this invention can be used to guide dosing regimens in further clinical trials (Phase I, II, or III).

[0079] Body weight and total protein significantly affected the drug's CL (closing threshold). The effects were smaller for body weight (45 kg and 90 kg, with relative median changes of -11.6% and 12.7%, respectively), while the effects of total protein (relative median changes of 6.8% and -6.3%, respectively) were not expected to be clinically significant. The model-estimated CL was also significantly affected by different blood sampling sites. Age significantly affected the drug's V1 (volumetric threshold). In the trial, results showed no effect on AUC0-24 in the age range of 19.8–53 years, but an effect on Cmax, which was not expected to be clinically significant. Attached Figure Description

[0080] Figure 1 This is a structural diagram of the three-chamber model of the present invention; Figure 2 This is the final diagnostic diagram of the drug model in Example 1 of the present invention; Figure 3 This is a distribution diagram of inter-individual variation (ETA) of the final drug model in Example 1 of the present invention; Figure 4 This is a visual prediction test plot (pcVPC) of the drug prediction value correction in Example 1 of the present invention. Figure 5 This is the final diagnostic diagram of the drug model in Example 2 of the present invention; Figure 6 This is a visual prediction test plot (pcVPC) of the drug prediction value correction in Example 2 of the present invention. Figure 7 This is a drug blood concentration-time relationship graph from Example 1 of the present invention; Figure 8 This is a graph showing the drug blood concentration-time relationship in Example 2 of the present invention. Detailed Implementation

[0081] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments, but the scope of protection of the present invention includes, but is not limited to, these.

[0082] Example 1

[0083] This embodiment provides a method for determining a population pharmacokinetic model of propofol derivative drugs, including: 1. Data Acquisition: Clinical Trials Propofol derivative SAD: A placebo- and positive-drug-controlled, single-dose intravenous escalation trial in healthy Australian subjects; Propofol derivative SAD_02: A phase I escalation trial with a positive drug control and a single intravenous injection in healthy subjects in Australia; Propofol derivative SAD_03: Safety and tolerability trial of a single intravenous injection followed by 30 minutes of continuous intravenous infusion in Australian subjects; Propofol Derivative-101: A single-center, open-label, uncontrolled phase I escalation trial evaluating a single intravenous injection of propofol derivative emulsion in healthy Chinese subjects; Propofol Derivative-103: A single-center, open-label, randomized, two-stage, crossover study evaluating the interaction (DDI) between propofol derivative emulsion injection and rifampicin capsules in healthy Chinese subjects; Propofol Derivative-202: A multicenter, open-label, non-randomized, positive-controlled, dose-escalation phase IIa clinical study evaluating the tolerability, efficacy, and safety of propofol derivative emulsion injection for induction of general anesthesia in subjects undergoing elective surgery. Propofol Derivative-302: This is a phase III, multicenter, randomized, double-blind, parallel-controlled study of propofol to evaluate the efficacy and safety of propofol derivatives compared to propofol in the induction of general anesthesia in Chinese patients undergoing elective surgery.

[0084] 2. Determine the data to be included in the analysis: The sampling scheme is shown in Table 1 below: Table 1. Sampling Plan

[0085] The original dataset includes 2,609 measurable blood drug concentration data from 219 subjects (81 of whom had both arterial and venous blood sampling; when a subject received a different blood sampling method, they were considered a new individual, i.e., 300 individuals were included in the population analysis).

[0086] The included clinical trial data were evaluated to determine the pharmacokinetic dataset to be included in the analysis: Blood drug concentration data (see) Figure 7 ); Baseline demographic data: race, age, height, weight, and sex; Blood biochemical indicators: total blood protein content, creatinine clearance rate, aspartate aminotransferase (AST) and alanine aminotransferase (ALT); Blood collection site.

[0087] 3. Data processing: The data from step 2 is processed, including: The determination and processing of observations below the detection limit include: the detection limit is determined by the detection instrument, observations below the detection limit are not used for population pharmacokinetic analysis, and when the proportion of observations below the detection limit is greater than 15%, the likelihood function method is used to examine the impact of observations below the detection limit on model fitting and modeling parameters. The identification and handling of outlier data includes: checking for outliers in the sample by examining the subject's dosing time and the corresponding plasma concentration curve, and excluding outliers; The identification and handling of outliers includes: identifying outliers based on residual analysis of the initial modeling results, and excluding outliers; Handling of missing covariates: If the missing rate of covariates for subjects is <15%, for continuous covariates, the median in the dataset is used to fill in the missing values; for categorical covariates, the most common category is used to fill in the missing values. If the missing rate of covariates for subjects is >15%, no values ​​are filled in, and subjects with complete covariate information are given Bayesian estimation to conduct exploratory analysis of the PK parameters.

[0088] 4. Establishment of the initial basic model of population pharmacokinetics: An initial basic model is established based on the data processed in step 3, including: (1) The inter-individual differences in PK parameters are described using the following formula: ; (2) The variability of the residuals can be described using the following formula: ; (3) Based on the blood drug concentration and time curves, the objective function of the two nested models was compared to determine whether the decrease was significant. The one-compartment, two-compartment, and three-compartment models were compared, and the model with the most significant decrease was selected. Figure 1 The three-compartment model shown is used as the initial basic model, where CL represents the central compartment clearance rate, V1 represents the central compartment distribution volume, V2 represents the peripheral compartment 1 distribution volume, V3 represents the peripheral compartment 2 distribution volume, Q2 represents the clearance rate between peripheral compartment 1 and the central compartment, Q3 represents the clearance rate between peripheral compartment 2 and the central compartment, R0 represents the infusion rate, and K represents the elimination rate constant.

[0089] 5. Establishment of the final basic model of population pharmacokinetics: The final base model is built based on the initial base model in step 4, including: (1) Based on clinical knowledge and drug action mechanisms, the covariates to be included in the assessment were determined: Baseline demographic covariates: baseline age, sex, weight, and race; Blood biochemical covariates: creatinine clearance, total protein, aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase, and total bilirubin; Blood collection site.

[0090] (2) Establish the final basic model of population pharmacokinetics based on the covariates included in the evaluation.

[0091] 6. Establishment of population pharmacokinetic model: Based on the final base model from step 5, and by incorporating selected covariates, a population pharmacokinetic model is established, including: (1) Covariate pre-screening: The following formula is used to analyze the correlation between continuous covariates and PK parameters: ; The following formula is used to analyze the correlation between categorical covariates and PK parameters:

[0092] (2) Use the forward and backward methods to finally screen the covariates and establish a population pharmacokinetic model: Forward method: Each covariate is added sequentially to the initial structural model of the final base model in step (5). Based on the log-likelihood ratio test, when one covariate is added, if the objective function value decreases by more than 6.63 based on the criterion of p<0.01, the newly added covariate is considered significant. Based on the initial structural model, the covariate with the most significant influence is added first to form an improved model. Then, the statistically significant covariates selected in the previous step are tested on the improved model. This process is repeated until no significant covariates can be found, and the full model is obtained. Backtracking method: The process of removing covariates one by one based on the full model. If the objective function value increases by more than 10.83 after removing one covariate, then the removed covariate is considered significant on the p<0.001 criterion.

[0093] In the non-MEM model, a stepwise forward addition method was used to investigate propofol derivatives in the PopPK model. The results showed that WT, TP, and SITE significantly affected CL (p<0.01), and AGE significantly affected V1 (p<0.01). Based on this, RACE was examined, and no significant effect was found on CL or V1. Using a stepwise backward elimination method, no influencing factors were eliminated (Δ...). -2LL >10.83). In the non-MEM model, a stepwise forward addition method was used for propofol PopPK, and the results showed that WT had a significant effect on CL (p<0.01). Using a stepwise backward elimination method, the effect of WT on the CL of propofol derivatives was eliminated (Δ-2LL=8.691).

[0094] (3) Model establishment

[0095] In the final PopPK model of propofol derivatives, body weight and total protein have a significant impact on CL, and the CL estimated by the model is also significantly affected by different blood collection sites; age has an impact on V1.

[0096] 7. Evaluation of population pharmacokinetic models: The pharmacokinetic model established in step 6 was evaluated, including: (1) Model Fit Diagnostic Chart (GOF) Diagnostic diagram of the final PopPK model of the drug as follows Figure 2 As shown, the results indicate a good consistency between the observed and predicted concentrations, with no significant deviation observed in the conditional weighted residual plots of the predicted concentrations over time. Figure 2The top left plot is a diagnostic plot of the observed values ​​and individual predicted values ​​of the final drug model; the top right plot is a diagnostic plot of the observed values ​​and population predicted values ​​of the final drug model; the bottom left plot is a diagnostic plot of the conditional weight residuals (CWRES) and time; the bottom right plot is a diagnostic plot of CWRES and population predicted values. The solid line in the middle is the standard line of 0, and the dashed line is the auxiliary line for |CWRES|=5.

[0097] (2) The distribution of the inter-individual random effects (ETA) in the final model of drug PopPK is shown in [reference needed]. Figure 3 As shown. The results indicate that the ETA is basically symmetrically distributed around 0. Among them, Figure 3 The vertical axis represents frequency, etaCL represents the inter-individual variation of CL, etaV1 represents the inter-individual variation of V1, etaQ2 represents the inter-individual variation of Q2, etaQ3 represents the inter-individual variation of Q3, etaV3 represents the inter-individual variation of V3, and the bold line in the middle is the standard line of 0.

[0098] (3) Visualized prediction test for predicted value correction (pcVPC)

[0099] pcVPC is used to evaluate the model's ability to reproduce data distributions. Using observed covariate information for each subject, final estimates of population pharmacokinetic model parameters, random effects, and residual errors, 1000 trials were simulated and reproduced. See pcVPC for drug plasma concentration-time curves. Figure 4 As shown, Figure 4 The data shows the observed concentrations (points) for all subjects, the 95% confidence interval (middle shaded area) of the model-predicted median, and the 95% confidence intervals (upper and lower shaded areas) of the model-predicted 2.5th and 97.5th quantiles. The results indicate that the final population pharmacokinetic model can adequately predict the central tendency and variability of drug concentrations in all subjects during the clinical study. Figure 4 The circles represent the observed values, the solid line in the middle represents the median of the observed values, the upper and lower dashed lines correspond to the 97.5th and 2.5th quantiles respectively, the shaded area in the middle represents the 95% confidence interval of the median predicted by the model, and the upper and lower shaded areas represent the 95% confidence intervals of the median and the 2.5th and 97.5th quantile lines of 1000 simulation data. Data below the detection limit are not shown in this figure.

[0100] (4) Bootstrap

[0101] The estimated parameters of the final model are compared with those of the Bootstrap model in Table 2 below. The median of the Bootstrap parameter estimates is similar to that of the final PopPK model. The 95% CI of the final PopPK model estimates (represented by the 2.5th to 97.5th quantile interval) highly overlaps with that of the Bootstrap model estimates, indicating that the final model has good stability and accuracy.

[0102] Table 2. Final model estimation parameters and Bootstrap estimation parameters

[0103] (4) Shrinkage

[0104] The final shrinkage values ​​of the model parameters are shown in Table 3.

[0105] In the final population pharmacokinetic model, ω V1 ω Q2 ω Q3 ω V3 The contraction was slightly higher than 30%.

[0106] Table 3. Final model parameter shrinkage values

[0107] in conclusion: The three-compartment linear pharmacokinetic model can well describe the pharmacokinetic characteristics of propofol derivatives within the dosage range of this study. Model evaluation shows that the final model has good stability, accuracy, and predictive performance.

[0108] Body weight and total protein significantly affected the CL (closing factor) of propofol derivatives, with smaller effects in the body weight ranges of 45 kg and 90 kg (relative median changes of -11.6% and 12.7%, respectively), and the effect of total protein (relative median changes of 6.8% and -6.3%, respectively) was not expected to be clinically significant. The model-estimated CL was also significantly affected by different blood collection sites; age significantly affected the V1 (volumetric quotient) of propofol derivatives, with results in the propofol derivative-302 trial showing a significant impact on AUC within the age range of 19.8–53 years. 0-24 It has no effect, but it does affect Cmax, though this is not expected to have clinical significance.

[0109] Example 2

[0110] This embodiment provides a method for determining a population pharmacokinetic model of propofol, including: 1. Data Acquisition: Clinical Trials A phase III, multicenter, randomized, double-blind, parallel-controlled study of propofol was conducted to evaluate the efficacy and safety of propofol in induction of general anesthesia in Chinese patients undergoing elective surgery.

[0111] 2. Determine the data to be included in the analysis: The sampling scheme is shown in Table 4 below: Table 4. Propofol Sampling Procedure

[0112] The propofol pharmacokinetic analysis dataset included 82 measurable blood concentration data from 28 subjects (60 of the 88 propofol subjects in the trial were not included in the analysis because they received propofol treatment for other purposes after induction anesthesia).

[0113] The included clinical trial data were evaluated to determine the pharmacokinetic dataset to be included in the analysis: Blood drug concentration data: obtained according to the implementation of the sampling protocol (see...). Figure 8 ); Baseline demographic data: race, age, height, weight, and sex; Blood biochemical indicators: total blood protein content, creatinine clearance rate, aspartate aminotransferase (AST) and alanine aminotransferase (ALT); Blood collection site.

[0114] 3. Data processing: The data from step 2 is processed, including: The determination and processing of observations below the detection limit include: the detection limit is determined by the detection instrument, observations below the detection limit are not used for population pharmacokinetic analysis, and when the proportion of observations below the detection limit is greater than 15%, the likelihood function method is used to examine the impact of observations below the detection limit on model fitting and modeling parameters. The identification and handling of outlier data includes: checking for outliers in the sample by examining the subject's dosing time and the corresponding plasma concentration curve, and excluding outliers; The identification and handling of outliers includes: identifying outliers based on residual analysis of the initial modeling results, and excluding outliers; Handling of missing covariates: If the missing rate of covariates for subjects is <15%, for continuous covariates, the median in the dataset is used to fill in the missing values; for categorical covariates, the most common category is used to fill in the missing values. If the missing rate of covariates for subjects is >15%, no values ​​are filled in, and subjects with complete covariate information are given Bayesian estimation to conduct exploratory analysis of the PK parameters.

[0115] 4. Establishment of the initial basic model of population pharmacokinetics: An initial basic model is established based on the data processed in step 3, including: (1) The inter-individual differences in PK parameters are described using the following formula: ; (2) The variability of the residuals can be described using the following formula: ; (3) Based on the blood drug concentration and time curves, the objective function of the two nested models was compared to determine whether the decrease was significant. The one-compartment, two-compartment, and three-compartment models were compared, and the model with the most significant decrease was selected. Figure 1 The three-compartment model shown is used as the initial basic model, where CL represents the central compartment clearance rate, V1 represents the central compartment distribution volume, V2 represents the peripheral compartment 1 distribution volume, V3 represents the peripheral compartment 2 distribution volume, Q2 represents the clearance rate between peripheral compartment 1 and the central compartment, Q3 represents the clearance rate between peripheral compartment 2 and the central compartment, R0 represents the infusion rate, and K represents the elimination rate constant.

[0116] 5. Establishment of the final basic model of population pharmacokinetics: The final base model is built based on the initial base model in step 4, including: (1) Based on clinical knowledge and drug action mechanisms, the covariates to be included in the assessment were determined: Baseline demographic covariates: baseline age, sex, weight, and race; Blood biochemical covariates: creatinine clearance, total protein, aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase, and total bilirubin; Blood collection site.

[0117] (2) Establish the final basic model of population pharmacokinetics based on the covariates included in the evaluation.

[0118] 6. Establishment of population pharmacokinetic model: Based on the final base model from step 5, and by incorporating selected covariates, a population pharmacokinetic model is established, including: (1) Covariate pre-screening: The following formula is used to analyze the correlation between continuous covariates and PK parameters: ; The following formula is used to analyze the correlation between categorical covariates and PK parameters:

[0119] (2) Use the forward and backward methods to finally screen the covariates and establish a population pharmacokinetic model: Forward method: Each covariate is added sequentially to the initial structural model of the final base model in step (5). Based on the log-likelihood ratio test, when one covariate is added, if the objective function value decreases by more than 6.63 based on the criterion of p<0.01, the newly added covariate is considered significant. Based on the initial structural model, the covariate with the most significant influence is added first to form an improved model. Then, the statistically significant covariates selected in the previous step are tested on the improved model. This process is repeated until no significant covariates can be found, and the full model is obtained. Backtracking method: The process of removing covariates one by one based on the full model. If the objective function value increases by more than 10.83 after removing one covariate, then the removed covariate is considered significant on the p<0.001 criterion.

[0120] In the NONMEM model of propofol, a stepwise forward addition method was used, and the results showed that WT had a significant effect on CL (p < 0.01). Using a stepwise backward elimination method, the effect of WT on propofol's CL was eliminated (Δ...). -2LL =8.691)

[0121] No significant effect of covariates on the pharmacokinetic (PK) parameters of propofol was found during the covariate screening.

[0122] 7. Evaluation of population pharmacokinetic models: The pharmacokinetic model established in step 6 was evaluated, including: (1) Model Fit Diagnostic Chart (GOF) Diagnostic diagram of the final PopPK model of the drug as follows Figure 5 As shown, the results indicate a good consistency between the observed and predicted concentrations, with no significant deviation observed in the conditional weighted residual plots of the predicted concentrations over time. Figure 5 The top left plot is a diagnostic plot of the observed values ​​and individual predicted values ​​of the final drug model; the top right plot is a diagnostic plot of the observed values ​​and population predicted values ​​of the final drug model; the bottom left plot is a diagnostic plot of the conditional weight residuals (CWRES) and time; the bottom right plot is a diagnostic plot of CWRES and population predicted values. The solid line in the middle is the standard line of 0, and the dashed line is the auxiliary line for |CWRES|=5.

[0123] (2) Visualized prediction test for predicted value correction (pcVPC)

[0124] pcVPC is used to evaluate the model's ability to reproduce data distributions. Using observed covariate information for each subject, final estimates of population pharmacokinetic model parameters, random effects, and residual errors, 1000 trials were simulated and reproduced. See pcVPC for drug plasma concentration-time curves. Figure 6 As shown, Figure 6The data shows the observed concentrations (points) for all subjects, the 95% confidence interval (middle shaded area) of the model-predicted median, and the 95% confidence intervals (upper and lower shaded areas) of the model-predicted 2.5th and 97.5th quantiles. The results indicate that the final population pharmacokinetic model can adequately predict the central tendency and variability of drug concentrations in all subjects during the clinical study. Figure 6 The circles represent the observed values, the solid line in the middle represents the median of the observed values, the upper and lower dashed lines correspond to the 97.5th and 2.5th quantiles respectively, the shaded area in the middle represents the 95% confidence interval of the median predicted by the model, and the upper and lower shaded areas represent the 95% confidence intervals of the median and the 2.5th and 97.5th quantile lines of 1000 simulation data. Data below the detection limit are not shown in this figure.

[0125] (3) Bootstrap

[0126] The estimated parameters of the final model are compared with those of the Bootstrap model in Table 5 below. The median of the Bootstrap parameter estimates is similar to that of the final PopPK model. The 95% CI of the final PopPK model estimates (represented by the 2.5th to 97.5th quantile interval) highly overlaps with that of the Bootstrap model estimates, indicating that the final model has good stability and accuracy.

[0127] Table 5. Final model estimation parameters and Bootstrap estimation parameters

[0128] (4) Shrinkage

[0129] The shrinkage values ​​of the final model parameters for propofol are shown in Table 6. Circles represent observed values, solid lines represent the median of observed values, and the upper and lower dashed lines correspond to the 97.5th and 2.5th quantiles, respectively. The shaded area marked 1 represents the 95% confidence interval of the predicted median, and the shaded area marked 2 represents the 95% confidence interval of the median and the 2.5th and 97.5th quantile lines from 1000 simulations. Data below the detection limit are not shown in this figure. In the final population pharmacokinetic model of propofol, ω... V1 The contraction was slightly higher than 30%.

[0130] Table 6. Shrinkage values ​​of the final model parameters for propofol

[0131] in conclusion: The three-compartment linear pharmacokinetic model can well describe the pharmacokinetic characteristics of propofol within the dosage range in this study. Model evaluation shows that the final model has good stability, accuracy, and predictive performance.

[0132] No covariates were found to have a significant effect on the pharmacokinetic parameters of propofol.

[0133] Example 3

[0134] This embodiment provides a system for determining the clinical individual dosing parameters of a compound of formula (I).

[0135] The system includes a data acquisition device, a data processing device, and a result output device; The data includes baseline demographic data, blood biochemical parameters, and blood sampling sites; the data processing device includes the ability to obtain individual dosing parameters (dosage) of compound (I) using the following formula: ;

[0136] CL i SITE represents the central compartment clearance rate of the i-th subject; SITE=0 when the blood is drawn from a vein and SITE=1 when the blood is drawn from an artery; WT represents body weight; TP represents total protein; η CL,i Let η be the inter-individual variation of CL for the i-th subject, and let η follow a normal distribution with mean 0 and variance ω2, where ω2 is the element on the diagonal of the variance-covariance matrix Ω of the inter-individual variation. V 1i This represents the central ventricular distribution volume of the i-th subject. V 2i This represents the peripheral 1 distribution volume of the i-th subject. V 3i This represents the peripheral 2 distribution volume of the i-th subject. Q 2i This represents the interventricular clearance rate between the peripheral chamber 1 and the central chamber for the i-th subject. Q 3i Let represent the peripheral ventricular clearance rate and the central ventricular clearance rate of the i-th subject, respectively. AGE represents age; η represents the inter-individual variation of the corresponding parameter.

[0137] Example 4

[0138] This embodiment provides a system for individual propofol dosing parameters. The system includes a data acquisition device, a data processing device, and a result output device. The data processing device includes obtaining the individual propofol dosing parameter (dosage) results using the following formula:

[0139] Among them, CL i η represents the central ventricular clearance rate of the i-th subject; CL,i Let η be the inter-individual variation of CL for the i-th subject, and let η follow a normal distribution with mean 0 and variance ω2, where ω2 is the element on the diagonal of the variance-covariance matrix Ω of the inter-individual variation. V 1i This represents the central ventricular distribution volume of the i-th subject. V 2i This represents the peripheral 1 distribution volume of the i-th subject. V 3i This represents the peripheral 2 distribution volume of the i-th subject. Q 2i This represents the interventricular clearance rate between the peripheral chamber 1 and the central chamber for the i-th subject. Q 3i Let represent the peripheral ventricular clearance rate and the central ventricular clearance rate of the i-th subject, respectively. AGE represents age; η represents the inter-individual variation of the corresponding parameter.

Claims

1. A method for population pharmacokinetic modeling of a deterministic compound (I), wherein, The method includes determining the following formula as a population pharmacokinetic model for the compound of formula (I): Equation (I), Among them, the pharmacokinetic parameter formulas in the population pharmacokinetic model of compound (I) include: ； Among them, CL i SITE represents the central compartment clearance rate of the i-th subject; SITE=0 when the blood is drawn from a vein and SITE=1 when the blood is drawn from an artery; WT represents body weight; TP represents total protein; η CL,i Let η be the inter-individual variation of CL for the i-th subject, and let η follow a normal distribution with mean 0 and variance ω2, where ω2 is the element on the diagonal of the variance-covariance matrix Ω of the inter-individual variation. The pharmacokinetic parameter formulas in the population pharmacokinetic model of compound (I) further include: Among them, V 1i This represents the central ventricular distribution volume of the i-th subject. V 2i This represents the peripheral 1 distribution volume of the i-th subject. V 3i This represents the peripheral 2-distribution volume of the i-th subject. Q 2i This represents the interventricular clearance rate between the peripheral chamber 1 and the central chamber for the i-th subject. Q 3i Let represent the peripheral ventricular clearance rate and the central ventricular clearance rate of the i-th subject, respectively. AGE represents age; η represents the inter-individual variation of the corresponding parameter.

2. The method according to claim 1, wherein, The method includes the following steps: (1) Data acquisition; (2) Determine the data to be included in the analysis; (3) Data processing; (4) Establishment of the initial basic model of population pharmacokinetics; (5) Establishment of the final basic model of population pharmacokinetics; (6) Establishment of a population pharmacokinetic model; (7) Evaluation of population pharmacokinetic models.

3. The method according to claim 2, wherein, Step (1) Data source: clinical trial data.

4. The method according to claim 2, wherein, Step (2) involves determining the pharmacokinetic dataset to be included in the analysis by evaluating the included clinical trial data.

5. The method according to claim 4, wherein, The included clinical trial data includes blood drug concentration data, baseline demographic data, blood biochemical parameters data, and blood collection sites.

6. The method according to claim 5, wherein, The baseline demographic data includes any two or more combinations of race, age, height, weight, and sex; the blood biochemical data includes any two or more combinations of total blood protein content, creatinine clearance, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and total bilirubin.

7. The method according to claim 2, wherein, Step (3) includes the identification and processing of combined data of one or more of the following: observations below the detection limit, outlier data, outliers, and missing covariates.

8. The method according to claim 7, wherein, The determination and processing of observations below the detection limit include: the detection limit is determined by the detection instrument, observations below the detection limit are not used for population pharmacokinetic analysis, and when the proportion of observations below the detection limit is greater than 15%, the likelihood function method is used to examine the impact of observations below the detection limit on model fitting and modeling parameters. The identification and handling of outlier data includes: checking for outliers in the sample by examining the subject's dosing time and the corresponding plasma concentration curve, and excluding outliers; The identification and handling of outliers includes: identifying outliers based on residual analysis of the initial modeling results, and excluding outliers; Handling of missing covariates: If the missing rate of covariates for subjects is <15%, for continuous covariates, the median in the dataset is used to fill in the missing values; for categorical covariates, the most common category is used to fill in the missing values. If the missing rate of covariates for subjects is >15%, no values ​​are filled in, and subjects with complete covariate information are given Bayesian estimation to conduct exploratory analysis of the PK parameters.

9. The method according to claim 8, wherein, The outlier data includes: 1) There are repeated concentration records at the same time point, except for arterial and venous concentrations at the same time point; 2) The valley concentration is greater than the corresponding peak concentration; 3) The administration time occurs after the peak concentration; 4) The administration time occurs before the trough concentration; 5) Concentration records after intravenous administration, prior to peak concentration; 6) Unexplained sudden drops or rises in concentration.

10. The method according to claim 2, wherein, Step (4) involves comparing multiple structural models based on blood drug concentration and time curves, selecting the best one as the initial structural model, and combining it with the residual model to form the initial basic model.

11. The method according to claim 10, wherein, The initial structural model is a three-compartment model with zero-order absorption and first-order linear elimination in the central compartment. The parameters of the initial structural model include: central compartment clearance rate CL, central compartment distribution volume V1, peripheral compartment 1 distribution volume V2, peripheral compartment 2 distribution volume V3, clearance rate between peripheral compartment 1 and the central compartment Q2, clearance rate between peripheral compartment 2 and the central compartment Q3, infusion rate R0, and elimination rate constant K.

12. The method according to claim 2, wherein, The initial basic model includes using the following formula to describe the inter-individual differences in PK parameters: in θi This represents the PK parameter of the i-th subject; θ T The natural logarithm of the typical values ​​of the PK parameter in the population; η i The variation among individuals is expressed as follows: the mean is 0 and the variance is ω. 2 A normally distributed random variable, where ω 2 The value represents the element on the diagonal of the variance-covariance matrix of the variation among individuals.

13. The method according to claim 2, wherein, The initial basic model includes the following formula to describe the variability of the residuals: in y ij This represents the j-th observed concentration of the i-th subject. ij ε represents the model-predicted concentration for the i-th subject. ij Let represent the proportional residual of the j-th observed concentration for the i-th subject. The observed concentration and the predicted concentration are independent of each other and follow normal distributions with mean 0 and variance σ², respectively.

14. The method according to claim 2, wherein, Step (5) includes determining the covariates to be included in the assessment based on clinical knowledge and drug action mechanisms, and establishing the final basic model of population pharmacokinetics based on the covariates to be included in the assessment.

15. The method according to claim 14, wherein, The covariates included in the assessment include baseline demographic characteristics, blood biochemical parameters, and blood collection sites.

16. The method according to claim 15, wherein, The baseline demographic covariates include any two or more combinations of age, sex, weight, and ethnicity; the blood biochemical covariates include any two or more combinations of creatinine clearance, total protein, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and total bilirubin.

17. The method according to claim 2, wherein, Step (6) includes a) Pre-screening of covariates; b) Use forward and backward methods to finally screen the covariates and establish a population pharmacokinetic model.

18. The method according to claim 17, wherein, The covariate pre-screening includes: The correlation between PK parameters and various covariates was analyzed using graphical methods. Linear regression was used for continuous covariates, and analysis of variance was used for categorical covariates. Based on the model evaluation dataset, Bayesian methods were used to estimate the parameters of the subjects in the final baseline model and to estimate the impact of covariates on PK parameters.

19. The method according to claim 18, wherein, The pre-screening of covariates includes, The following formula is used to analyze the correlation between continuous covariates and PK parameters: ； The following formula is used to analyze the correlation between categorical covariates and PK parameters: in θ i This represents the PK parameter of the i-th subject; θ pop This represents the typical value of the PK parameter in the subject population; Cov i This represents the continuous covariate value for the i-th subject; Cov pop This represents the median of a continuous variable in the test group; X i Let represent the categorical variable index for the i-th subject, where a value of 0 indicates the category of the covariate with the most common category, and other integer values ​​indicate other categories; k cov This represents a coefficient that describes the magnitude of the influence of the covariate.

20. The method according to claim 19, wherein, The final screening of covariates includes: Based on the final basic model, a full model is built using the forward method, and then a population pharmacokinetic model is built using the backward method based on the full model. The forward method includes: adding each covariate sequentially to the initial structural model of the final base model in step (5). Based on the log-likelihood ratio test, when one covariate is added, if the objective function value decreases by more than 6.63 based on the criterion of p < 0.01, the newly added covariate is considered significant. Based on the initial structural model, the covariate with the most significant impact is added first to form an improved model. Then, the statistically significant covariates selected in the previous step are tested on the improved model. This process is repeated until no significant covariates can be found. The backtracking method includes: a process of deleting covariates one by one based on the full model. If the objective function value increases by more than 10.83 after deleting one covariate, then the deleted covariate is considered significant on the p<0.001 criterion.

21. The method according to claim 2, wherein, Step (7) includes evaluating the population pharmacokinetic model using one or more of the following methods in combination: model fit diagnostic plot, visual prediction test, bootstrapping and contraction method.

22. The method according to claim 21, wherein, The model fitting diagnostic plot includes one or more combinations of the following graphs: population predicted concentration vs. observed concentration, individual predicted concentration vs. observed concentration, conditionally weighted residual vs. population predicted concentration, and conditionally weighted residual vs. time after the first dose. Visualized prediction verification involves plotting and comparing the predicted results with the measured values ​​using the final model parameters, covariates, and actual doses to assess whether the population pharmacokinetic model can adequately describe the pharmacokinetic curve of compound (I). The bootstrapping method involves repeatedly fitting a population pharmacokinetic model to 1000 datasets for bootstrapping replication, and randomly selecting subject data and covariates for replacement to replicate the model. The contraction method involves estimating individual parameter values ​​of subjects using Bayesian estimation based on a population pharmacokinetic model, and calculating inter-individual variability and individual residuals from model predictions and observed values.

23. The method according to claim 22, wherein, The contraction method involves using the following formula to assess inter-individual variability and individual residuals of pharmacokinetic parameters, and to quantify individual parameter values ​​and random error estimates: ； in, η shrinkage For inter-individual variation, ε shrinkage Let ω represent the individual residuals, and ω represent the inter-individual variation of the individual parameter values ​​estimated by the population pharmacokinetic model. η ph This parameter is for all individuals η The values ​​are IWRES, which represent the individual weighted residuals, and SD, which represents the standard deviation.

24. The method according to claim 2, wherein, Step (7) also includes estimating the individual PK parameters of the subject using the Bayesian post-hoc method, simulating the drug-time curve of intravenous infusion based on the actual administered dose, and calculating the area under the drug-time curve for 0 to 1 minute, 0 to 2 minutes, 0 to 4 minutes, 0 to 10 minutes, 0 to 24 hours, and peak concentration.

25. A system for individual dosing parameters of a compound of formula (I), said system comprising a data acquisition device, a data processing device, and a result output device. Equation (I) in, The data includes baseline demographic data, blood biochemical index data, and blood collection site information data; the data processing device includes obtaining the individual dosing parameters of compound (I) using the pharmacokinetic parameter formula in the population pharmacokinetic model of claim 1.

Images