JP2018190244A - Novel prediction method of concentration transition in human plasma using physiological model - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method and a device for highly accurately predicting (simulating) a transition of a compound concentration in human plasma.SOLUTION: A prediction (simulation) method for predicting a transition of a compound concentration in human plasma by using a computer, comprises the steps of: setting predetermined parameters in a differential equation related to pharmacokinetics; and calculating a time transition of the compound concentration in human plasma by numerically analyzing the differential equation. The predetermined parameters include: a) a physical property value of a compound; b) a biological value of the compound related to a model animal and/or a human being; and c) a value calculated from at least one value among an in vivo measured value or a physical property value of the compound obtained from an in vivo kinetics trial using model animals, the biological value, and the in vivo measured value.SELECTED DRAWING: Figure 7

Description

  The present disclosure relates to a method and apparatus for predicting (simulating) the transition of a compound concentration in human plasma by a physiological model.

  In the field of drug discovery, it takes a very long time from discovery of a new substance as a seed of a pharmaceutical to confirmation of its effectiveness in humans. Until clinical trials in which drug candidates are administered to humans, efficacy and safety are valid as drug candidates by non-clinical tests represented by animals (in vivo) and in vitro experiments (in vitro). Is considered. In order to accelerate research and development of drug discovery and deliver high-quality pharmaceuticals to patients as soon as possible, it is extremely important to accurately predict the effectiveness and safety in humans from non-clinical studies. Effectiveness and safety depend on the amount of compound administered into the living body, so it is considered that determining the amount of compound in the human body with high accuracy directly leads to acceleration of drug development.

  The research area for analyzing quantitative changes in compounds in vivo over time is called pharmacokinetic (hereinafter “PK”). In general, PK can be generally understood by grasping ADME (A: absorption, D: distribution, M: metabolism, E: excretion). After the compound administered to animals and humans is absorbed from the digestive tract, etc., it is distributed to each tissue through the whole body blood, and after being converted into a form that is easily excreted by undergoing metabolism in the disappearing organ such as the liver, It is excreted ex vivo via feces or urine. In non-clinical studies, it is possible to determine the PK of a compound in laboratory animals, but there are species differences in blood flow velocity, tissue size, metabolic activity, etc. between humans and laboratory animals. The obtained PK cannot be used as a human PK as it is. Thus, there are methods for predicting human PK from experimental animals.

  Animal scale-up is widely recognized as a method for predicting human PK from experimental animals. In this method, after administering a compound to rats, dogs, or monkeys, plasma concentration transitions are evaluated over time, and human PK is empirically predicted using species differences such as body weight and brain weight as correction factors (Non-Patent Documents). 9, 10). This prediction method has high prediction accuracy for human PK when the metabolic activity in animals and humans is almost the same, and there is no species difference in the protein binding rate in plasma that plays a part in tissue distribution. On the other hand, with animal scale-up, there are issues such as the need for highly evolved dogs and monkeys, and if there are species differences in metabolic activity and plasma protein binding rate, the prediction accuracy of human PK decreases. Is recognized.

  In order to solve this problem, the concept of physiologically based pharmacokinetics (PBPK) has been established since the mid-1990s (Non-Patent Documents 9 and 10). This method generates on a computer a biological model based on physiological predispositions such as the main tissue of the living body, its constituent components and the blood flow velocity of the tissue, and inputs parameters such as metabolic activity and protein binding rate to this biological model. By doing so, PK can be predicted. When predicting human PK, metabolic activity using human hepatocytes, etc., protein binding rate in blood, etc. will be evaluated in an in vitro test and input to a biological model on a computer. Therefore, species differences such as metabolic activity between experimental animals and humans, such as animal scale-up, do not affect the prediction accuracy. In recent years, since this PBPK is expected to theoretically be able to predict PK and drug-drug interactions (DDI), the discussion is extremely active. In the International Consortium for Innovation & Quality in Pharmaceutical Development (IQ) PBPK working group, the goal of the difference between the actual measurement and prediction of human PK within two times was set (Non-patent Document 11). On the other hand, in the report on human PK prediction using PBPK, the prediction accuracy is verified using many compounds, but the target set by IQ PBPK working group is not fully achieved. There are two possible causes. One is a problem related to the prediction target such as racial and individual differences, and the other is a problem of PK prediction process. The former has reported a methodology that makes it possible to predict human PK by taking into account metabolic activity and the like, including racial and individual differences. The latter problem is that many mathematical assumptions are introduced in order to model a living body on a computer so that it can be predicted quickly and easily. Thus, in our view, the distribution volume of compounds in humans, the transferability from the blood to the liver, and the metabolic rate in the liver cannot be accurately predicted. From this, in the compound group which shows a specific property, the tendency which underestimates the disappearance rate of a compound at the time of human PK prediction by PBPK is recognized.

  The current PBPK human PK prediction is difficult to place high reliability on the prediction accuracy. Therefore, there are almost no cases where it is used as evidence material when applying for or consulting new pharmaceuticals. On the other hand, it is possible to predict the drug-drug interaction (DDI) with high accuracy by utilizing the clinical trial results and PBPK in an integrated manner. This analysis discusses the significance of the DDI study in the phase 2 study, and is used to accelerate clinical trials.

Journal of Pharmaceutical Science, 2005 Jun; Vol. 94 (6): 1259-76 Journal of Pharmaceutical Science, 2006 Jun; Vol. 95 (6): 1238-57 Pharmaceutical Research, 2007 May; Vol. 24 (5): 918-33 Journal of Pharmacokinet Biopharm, 1985 Oct; Vol. 13 (5): 477-92 Pharmaceutical Research, 2009 Aug; Vol.26 (8): 1881-9 Journal of Pharmaceutical Science, 2012 Feb; Vol.101 (2): 838-51 Journal of Pharmaceutical Science, 2013 Jul; Vol.102 (7): 2085-95 Journal of Pharmaceutical Science, 2013 Sep; Vol.102 (9): 3239-51 Journal of Clinical Pharmacology, 2013 Feb; Vol. 53 (2): 167-77. Journal of Clinical Pharmacology, 2013 Feb; Vol. 53 (2): 178-91. Clinical Pharmacology & Therapeutics, 2015 Mar; Vol. 97 (3): 247-62

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a method and apparatus that can predict (simulate) human PK with high accuracy. More specifically, the present invention provides a method and an apparatus that can accurately calculate a human plasma compound concentration transition.

  As a result of searching for the use of PBPK for the purpose of improving the prediction accuracy of human PK prediction in non-clinical studies, the present inventors have used various parameters obtained from in vitro tests and minimal in vivo tests. We have devised the following method that enables human PK prediction with high accuracy using biological models based on concepts that have not been reported to PBPK.

  In a first aspect of the present invention, a method is provided for predicting (simulating) a transition of a human plasma compound concentration using a computer. The prediction method includes a step of setting a predetermined parameter in a differential equation relating to pharmacokinetics, and a step of numerically analyzing the differential equation to calculate a time course of a human plasma compound concentration. The predetermined parameters are: a) a physical property value of the compound, b) a biological value of the compound relating to the model animal and / or human, and c) an in vivo measurement value of the compound obtained from a pharmacokinetic test using the model animal. Alternatively, a physical property value, a value calculated from at least one of the biological value and the in vivo measurement value are included.

  In a second aspect of the present invention, a simulation apparatus for executing the prediction method of the first aspect is provided.

  In the third aspect of the present invention, a program for causing a computer to execute the prediction method of the first aspect is provided.

  According to the method and apparatus of the present invention, the accuracy of human PK prediction can be improved, and the transition of human plasma compound concentration can be accurately calculated.

Diagram showing human biological model for PK prediction Figure explaining Kpu calculation formula Diagram showing physiological parameters used for Kpu calculation Diagram showing physiological parameters used for Kpu calculation Illustration for explaining compounds that migrate to the liver Diagram explaining the relationship of compound concentration to the disappearing organ (liver) Illustration explaining the relationship of compound concentration to non-disappearing organs Block diagram showing the configuration of the simulation device Flow chart showing simulation processing of human plasma compound concentration transition by simulation device The figure which shows an example of the input screen of the calculation conditions in a simulation apparatus Figure showing simulation results of time course of plasma concentration in vein based on new PBPK theory for various compounds The figure which shows the comparison of the fold change for the PK parameter which is predicted based on the new PBPK theory and the PK parameter which is predicted based on the conventional method The figure which shows the comparison of the fold change for the PK parameter which is predicted based on the new PBPK theory and the PK parameter which is predicted based on the conventional method

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings as appropriate.

<Definition of parameters>
The parameters used in the following disclosure are defined as follows.
pKa (acid): Acid dissociation constant of acidic compounds. A smaller pKa indicates a stronger acid.
pKa (base): Acid separation constant of basic compounds. A larger pKa indicates a stronger base.
logP: Partition coefficient indicating hydrophobicity. In particular, it means the distribution coefficient when n-octanol and water are used as the solvent.
pHBC: Hydrogen ion index in blood cells
pHIW: Hydrogen ion index in intracellular fluid
pHP: Hydrogen ion index in plasma
fEX: the proportion of the tissue occupied by extracellular fluid
fIW: the proportion of the tissue occupied by intracellular fluid
fNL: Percentage of volume occupied by neutral fat in tissues
fNP: the proportion of the tissue occupied by neutral phospholipids
[AP] T: concentration of acidic phospholipids in tissue T
[PR] T: Concentration of extracellular albumin or lipoprotein in tissue T
GFR: Glomerular filtration rate
X: Dose. Regardless of the administration method, for example, oral administration, intravenous administration, intramuscular administration and the like can be mentioned.
Rate _inf : Dosing rate for constant-rate intravenous administration (0 for rapid intravenous injection)
t: When used as a subscript, means tissue (r: kidney, m: muscle, gi: intestine, h: liver, a: artery, v: vein, ad: fat, s: skin, lu: lung) . When used in a differential operator, it means time.

Ct: Tissue compound concentration (= C, tissue)
Cr: Compound concentration in the kidney
Cm: Compound concentration in muscle
Cgi: Compound concentration in the intestine
Ch: Compound concentration in the liver
Ca: Compound concentration in arteries flowing into tissues
Cv: Compound concentration in veins exiting the tissue
Cad: Compound concentration in fat
Cs: Compound concentration in the skin
Clu: Compound concentration in the lung
Ce: Compound concentration in the extracellular fluid (= C, e)
CLint: Hepatic metabolic clearance (inherent clearance obtained from liver cell metabolic stability and liver microsomal metabolic stability assessment). The clearance is defined as a value obtained by dividing the velocity by the concentration. Intrinsic clearance refers to the true compound processing capacity of the tissue.
fu, b: Percentage of compounds in blood that are not bound to blood proteins (= fu, blood)
fu, t: Percentage of compounds that are not bound to protein in tissue among compounds present in tissue (= fu, tissue)
fu, liver: Percentage of compounds that are not bound to liver protein among the compounds present in the liver
fu, p: Percentage of compounds in plasma that are not bound to plasma proteins
fu, p-app: Percentage of non-ionic compounds that are not bound to proteins in plasma (blood).
Kpu: Ease of migration of compounds that are not bound to proteins into tissues (tissue migration)
Kp, t, calculation: Ease of migration to rat tissues (tissue migration) obtained from the formula in the table of Fig. 2A
Kp, t: Ease of compound transfer in rat tissue when Kp, t, calculation is corrected by the ratio of measured rat Vdss and Vdss, rat, calculation (tissue migration)
Kp, t, human: Ease of transfer of compounds to tissues in humans (tissue transferability)
Kp: Ease of migration of compounds to tissues (tissue migration)
Kp, r: tissue migration of compounds in the kidney
Kp, m: tissue migration of compounds in muscle
Kp, gi: Tissue migration of compounds in the intestinal tract
Kp, h: tissue migration of compounds in the liver
Kp, a: tissue migration of compounds in arteries
Kp, v: Tissue migration of compounds in veins
Kp, ad: Tissue migration of compounds in fat
Kp, s: tissue migration of compounds in the skin
Kp, lu: tissue migration of compounds in the lung
Q: Blood flow velocity in each tissue (= Qt)
Qr: Blood flow velocity in the kidney
Qm: Blood flow velocity in muscle
Qgi: Blood flow velocity in the intestinal tract
Qh: Blood flow velocity in the liver
Qa: Blood flow velocity in arteries
Qv: Blood flow velocity in the vein
Qad: Blood flow velocity in fat
Qs: Blood flow velocity in the skin
Qlu: Blood flow velocity in the lung
Vdss, rat, calculation: Volume of distribution in the steady state of the rat derived from the formula in the table. The volume of distribution refers to the volume (L / kg or mL / kg) determined when it is assumed that the compound is instantaneously distributed in each tissue at a concentration equal to that in plasma.
Vdss: The distribution volume of a compound when the tissue transition of the compound reaches a steady state in vivo.
V: Tissue volume (= Vt). Refers to the volume (size) of tissue t.
Vr: Kidney tissue volume
Vm: muscle tissue volume
Vgi: Intestinal tissue volume
Vh: liver tissue volume
Va: Arterial tissue volume

Vv: Venous tissue volume
Vad: Fat tissue volume
Vs: skin tissue volume
Vlu: Lung tissue volume

  Hereinafter, the novel PBPK theory devised by the present inventors will be described.

1. New PBPK theory
PBPK theory predicts the pharmacokinetics of a compound over time based on human physiological parameters (blood flow velocity, organ size, etc.) and compound characteristics (e.g., protein binding, ease of transition to tissues, etc.) It is a method to do. In general, as shown in FIG. 1, the whole living body is regarded as a drug processing system, and the inside of the living body is divided into several processing sections (compartment model). For example, the stomach and intestine are considered as an absorption compartment, the whole blood as a distribution compartment, the liver as a metabolic compartment, and the kidney as an excretory compartment. For each compartment, the rate of change in the amount of drug substance is described by a differential equation. Note that the compartment model shown in FIG. 1 is an example. The new PBPK theory described below can be applied not only to the model shown in FIG. 1 but also to various compartment models. For example, the new PBPK theory can be applied to a compartment model to which brain and bone compartments are added.

  Prediction of pharmacokinetics is performed by solving simultaneous differential equations described for each compartment. For this purpose, it is necessary to determine parameters to be substituted for each differential equation. There are roughly two types of parameters. One is a physiological parameter, a known value. The other is a compound-derived parameter, such as metabolism-specific clearance (CLint) and tissue transferability (Kp, t).

  The new PBPK theory devised by the inventors of the present application differs from the conventional PBPK theory in two points. The first difference is a method for predicting tissue migration in the distribution process, and the second difference is the type of compound that undergoes liver migration and undergoes metabolism in the elimination process. Hereinafter, each difference will be described.

1.1 Distribution Process In the conventional prediction method, the human Kp, t of a compound is predicted from the value derived from the compound and human tissue components. The likelihood (predictive accuracy) of Kp, t can be verified based on the value of the distribution volume (Vdss) in the steady state that can be calculated based on Kp, t. In the conventional prediction method, the actual measurement and the predicted Vdss value may be different. Therefore, the present inventor has devised a predictable method with higher accuracy.

  Specifically, instead of directly predicting human Kp, t as in the prior art, first, rat Kp, t is predicted (step 1), rat Vdss is obtained from this value (step 2), and then Taking into account the species difference between rats and humans, human Kp, t is obtained (step 3). The details are shown below.

(Step 1)
Rat Kp, t is predicted as follows. First, for each tissue (t) of a rat, a calculated value (Kp, t, calculation) of rat Kp, t is obtained from the physical property value of the compound and vitro data according to the following formula (see FIGS. 2A to 2C).
Kp, t, calculation = Kpu * fu, plasma (1.1)

(Step 2)
Next, the calculated value (Vdss, rat, calculation) of rat Vdss is calculated by the following equation.
Vdss, rat, calculation = Σ (Kp, t, calculation * Vt) + blood vessel volume (1.2)

(Step 3)
By correcting the calculated value Kp, t, calculation of the rat Kp, t as described above based on the relationship (ratio) between the calculated value of the rat Vdss and the measured value (Vdss, rat), the rat Kp , t, the predicted value of Kp, t, rat is obtained (see the following formula).
Kp, t, rat = Vdss, rat / Vdss, rat, calculation * Kp, t, calculation (1.3)

  As described above, rat Kp, t, rat is obtained. Then, from the rat Kp, t, rat, Kp, t, human, which is a human Kp, t, is obtained in consideration of the species difference between human and rat.

Specifically, from the rat Kp, t, rat obtained as described above and the measured values of Vt, rat, Vt, human, fu, b, rat and fu, b, human, based on the following relationship, Obtain Kp, t, human.
C, t * fu, t = C, e * fu, b (1.4)
The following equation is obtained from equation (1.4).
Kp = C, t / Ce = fu, b / fu, t (1.5)

From the equation (1.5), the following equation is obtained.
fu, t = fu, b / Kp (1.6)

Here, the drug concentration of each tissue has the following correlation between human and rat.
Vt, human / fu, t, human = 1.0 * (Vt, rat / fu, t, rat) ^0.951 ... (1.7)

From the equations (1.6) and (1.7), the following equation is obtained.

  As described above, human Kp, t (Kp, t, human) is obtained from rat Kp, t (Kp, t, rat) in consideration of the species difference between human and rat. By the above method, there is no case where the prediction accuracy of Vdss is poor as in the conventional prediction method, and the human Kp, t obtained in this way is divided by Rb and input to the simultaneous differential equations of PBPK. It has become possible to predict changes in human plasma concentration with prediction accuracy. In formulas (1.7) and (1.8), the exponent value is set to 0.951 as an example, but the exponent value is not limited to 0.951. Depending on the theory, other values are possible.

1.2 Disappearance process The main organ in which a compound undergoes metabolism is the liver, and in order to undergo metabolism, the compound needs to be transferred from the blood to the liver. In the conventional PBPK theory, when there is no contribution of a transporter expressed in hepatocytes, only the membrane permeation by passive diffusion is considered for the compound transfer from the blood to the liver. However, this concept does not fully explain the behavior of the compound in the living body, and as a result, there are cases where the prediction result of the human plasma concentration differs from the actual measurement.

  Therefore, the inventors of the present application paid attention to the fact that compounds that bind to albumin present in blood also migrate to the liver (albumin-mediated uptake) and applied it to the novel PBPK theory. This makes it possible to predict the disappearance phase (disappearance process) with high accuracy, especially when predicting the plasma concentration transition.

  Conventionally, many model calculation formulas have been reported in which compounds that are not bound to protein (non-protein-binding compounds) among compounds existing in the liver are used as compounds that undergo drug metabolism. However, the calculation result of the human plasma concentration sometimes differs greatly from the actual measurement value.

  In contrast, the inventors of the present application have included non-ionic compounds in the model calculation formula in addition to non-protein-binding compounds as compounds that undergo metabolism in the liver. Thereby, the accuracy of the calculation result of the human plasma concentration could be improved.

  Specifically, albumin-binding compounds and α1 acidic glycoprotein-binding compounds are also included as compounds that migrate to the liver and undergo metabolism in the liver (see FIG. 3).

  Hereinafter, the derivation of the metabolic disappearance process for the albumin-binding compound and the α1 acidic glycoprotein-binding compound will be described.

(1) Derivation of metabolism disappearance process of albumin binding compound The compound disappearance rate (velim) from the tissue is represented by the following formula (see FIG. 4).
veilm = Q * (Ca-Cv) = C * fu, b * CLint (2.1)
The rate of compound disappearance from the tissue over time is represented by the following formula.
Vt * dCt / dt = Q * (Ca-Ce)-velim
= Q * (Ca-Ce)-Cv * fu, b * CLint (2.2)
Here, from Kp = Ct / Ce, Ce = Cv, the following equation is obtained from equation (1).
Vt * dCt / dt = Q * (Ca-Ct / Kp)-(Ct / Kp) * fu, b * CLint (2.3)

In the case of passive diffusion, the following equation holds for the tissue compound concentration.
Ce * fu, b = Ct * fu, t (2.4)

Therefore, the following equation holds.
Ct / Ce = fu, b / fu, t (= Kp) (2.5)
Substituting equation (2.5) into the last term of equation (2.3) yields:
Q * (Ca-Ct / Kp)-Ct * (fu, t / fu, b) * fu, b * CLint
= Q * (Ca-Ct / Kp)-Ct * fu, t * CLint (2.6)
Considering the liver, the following equation is obtained.
Vliver * dCt / dt = Q * (Ca-Ch / Kp)-Ch * fu, liver * CLint (2.7)

Therefore, velim becomes the following equation. The velim of the metabolic loss process of albumin-binding compounds is as follows.
velim = fu, liver * CLint * Ch ・ ・ ・ (2.8)

Here, fu and liver in the above equation are obtained based on the following equation.
fu, liver = PLR * fu, p-app / {1 + (PLR-1) * fu, p-app} ・ ・ ・ (2.9)

(2) Derivation of metabolic disappearance process of α1 acidic glycoprotein-binding compound As described above, the following equation holds.
velim = Ch * fu, liver * CLint ・ ・ ・ (2.10)
In the case of compound transfer from plasma to tissue, when passive diffusion is achieved with plasma (blood) non-ionized compounds, the following equation holds from free theory.
Cv * fu, p-app = Ct * fu, t ・ ・ ・ (2.11)
Ct / Cv = fu, p-app / fu, t ... (2.12)

Here, since Kp = Ct / Cv, Kp = fu, p-app / fu, t.
Therefore, the following equation is obtained.
fu, t = fu, p-app / Kp (2.13)
Substituting equation (2.13) into equation (2.10), we obtain the following equation for the disappearance rate of the α1 acidic glycoprotein-binding compound.
velim = fu, p-app * CLint * Ch / Kp, h (2.14)

(3) Basic formula of PBPK in non-disappearing organ The amount of compound distributed in the tissue (t) per unit time is represented by the following formula (see FIG. 5).
Vt * Ct = Q * (Ca-Ce) (3.1)

Since Kp = Ct / Cv = Ct / Ce, the following equation is obtained.
Ce = Ct / Kp (3.2)
Therefore, the following equation is obtained by substituting equation (3.2) into equation (3.1).
Vt * Ct = Q * (Ca-Ct / Kp, t) (3.3)
Differentiate the above equation by time t to obtain the following equation.
Vt * dCt / dt = Q * (Ca-Ct / Kp, t) (3.4)

  The PBPK formula for non-eliminated organs is basically obtained based on the above formula (3.4).

1.3 Simultaneous Differential Equations for Pharmacokinetics Based on the above concept, the following equations are obtained as simultaneous differential equations for pharmacokinetics for the compartment model shown in FIG.

  The above equation (4.4) is a differential equation relating to metabolic kinetics, and is an equation for calculating the time change of the human liver compound concentration. Formula (4.4) has a term of disappearance rate (velim) including the proportion of non-protein-bound compound in the liver (fu, liver, fu, p-app). The elimination rate velim is obtained from the formula (2.8) or the formula (2.14), and takes into consideration a protein non-binding compound and a nonionic compound as compounds that undergo metabolism in the liver.

  In the PK prediction method of the present embodiment based on the new PBPK theory, when numerically analyzing the above simultaneous differential equations, first, the human Kp, t is calculated from the rat Kp, t, and the value is calculated as a simultaneous differential equation (4.1). Assign (set) to ~ (4.9). Then, the simultaneous differential equations are numerically analyzed using a computer, and the transition of the compound concentration in each human tissue (plasma) is calculated.

  According to the new PBPK theory explained above, in the dispersion process, human Kp, t is predicted from the Kp, t of an animal (eg, rat) in consideration of the species difference of the animal, and its value is used to change human plasma concentration. Predict. In addition, as a compound that undergoes metabolism in the liver during the disappearance process, in addition to the protein non-binding compound, a nonionic compound was also included in the model calculation formula. Thereby, it became possible to predict the plasma concentration transition with high accuracy.

2. PBPK Simulation 2.1 Simulation Device Hereinafter, a simulation device that performs the PK prediction method based on the above-described novel PBPK theory and predicts (simulates) the transition of the compound concentration in each human tissue (in plasma) will be described.

  FIG. 6 shows a configuration example of the simulation apparatus. The simulation apparatus 10 is configured by an information processing apparatus such as a personal computer. The simulation apparatus 10 numerically analyzes the simultaneous differential equations of the above formulas (4.1) to (4.9), and calculates the transition of the compound concentration (C, tissue) in each human tissue (tissue).

  The simulation apparatus 10 includes a control unit 11 that controls the overall operation, a display unit 17 that performs screen display, an operation unit 19 that is operated by a user, and a data storage unit 21 that stores data and programs. For example, the display unit 17 is configured by a liquid crystal display, and the operation unit 19 is a keyboard, a mouse, or the like. Furthermore, the simulation apparatus 10 includes an interface 25 for connecting to an external device or a network. The simulation apparatus 10 can be realized by a personal computer, for example.

  The control unit 11 includes a CPU or MPU that executes a general-purpose program and a RAM. The control part 11 implement | achieves the function of PBPK simulation by running this program. This program is stored in the data storage unit 21 in advance. The program may be provided to the simulation apparatus 10 from the outside through a communication line or via a recording medium such as an optical disk or a semiconductor memory. The control unit 11 may be configured by a dedicated hardware circuit designed to realize a predetermined function.

  The data storage unit 21 is a means for storing data and programs, and can be composed of, for example, a hard disk (HDD), SSD, semiconductor memory, or the like. The data storage unit 21 stores programs executed by the control unit 11 and simulation result data. The data storage unit 21 further stores information related to a pharmacokinetic model for simulation (simultaneous differential equations, parameters) and information related to mathematical formulas (expressions (1.1) to (3.4)) for calculating various parameters.

2.2 Simulation Operation Hereinafter, a PK prediction operation (PBPK simulation) based on the new PBPK theory by the simulation apparatus 10 will be described. FIG. 7 is a flowchart showing a PBPK simulation process executed by the simulation apparatus 10 (control unit 11). By this PBPK simulation, the compound concentration transition in human plasma is determined.

  In FIG. 7, the simulation apparatus 10 (control unit 11) first inputs various information for PK prediction (S10). The simulation apparatus 10 displays an input screen for inputting information necessary for PK prediction on the display unit 17. The simulation apparatus 10 reads information set on the input screen.

  FIG. 8 shows an example of an input screen for inputting various information for PK prediction. As shown in FIG. 8, the input screen 60 includes input areas 61 to 65 for inputting information. The input area 61 is an area for designating a target animal species for PK prediction. In this example, “human” or “rat” can be selected (“human” is selected in FIG. 6). The input area 62 is an area for inputting values relating to the physical properties of the compound. Here, the actual measurement value is input, but if the prediction accuracy is high, the prediction value may be input. The input area 63 is an area for inputting metabolism-related parameters. The input area 64 is an area for inputting an in vivo measurement value of a compound obtained from a pharmacokinetic test of a human or a rat (an example of a model animal). The input area 65 is an area for inputting other parameters (for example, drug dosage (Does)) necessary for PK prediction.

  Next, the simulation apparatus 10 calculates parameters necessary for PK prediction based on the parameters input from the input screen 60 (S20). Specifically, the simulation apparatus 10 calculates parameters (CLint, fuinc, fuliver, fup-app, Kp, t) necessary for PK prediction using the parameters input from the input screen 60 and the above-described mathematical formulas. To do. In the example of FIG. 8, “human” is designated as the target animal species, and fuinc, fu, liver, fup-app, Kp, t for humans are calculated.

For example, CLint, fuinc, fu, liver, and fup-app parameters are calculated using values input in the input areas 61 to 64. The calculation method of these parameters is disclosed in the following document, for example.
Journal of Pharmaceutical Science, 2011 Oct; Vol. 100 (10): 4501-17
Journal of Pharmaceutical Science, 2012 Feb; Vol. 101 (2): 838-51
Journal of Pharmaceutical Science, 2013 Aug; Vol. 102 (8): 2806-18

Further, pKa and logP (that is, P) input to the input area 62 are used for calculating Kpu when obtaining human Kp, t from rat Kp, t. Furthermore, pKa and logP are also used for calculating fuinc. Fu input in the input area 63, plasma and R _BP is used in order to calculate the appropriate entries in Velim, fu input in the input area 64, plasma and R _BP is to calculate the rat Kpu It is used for.

  Since human and animal-specific physiological parameters such as Vt (tissue volume) and Qt (blood flow velocity) (see FIGS. 2B and 2C) are known values, they are stored in advance in the data storage unit 21 of the simulation apparatus 10. ing. Other parameters necessary for PK prediction are also stored in the data storage unit 21 in advance.

  In step S20, the simulation apparatus 10 obtains rat Kp, t and Kpu based on the input parameters (formulas (1.1) to (1.3), see FIG. 2A), and calculates human Kp, t using these values. (See equation (1.4)).

  Next, the simulation apparatus 10 executes a PBPK simulation using the calculated parameters (S30).

  In the PBPK simulation, parameters are set in the simultaneous differential equations of equations (4.1) to (4.9), and the simulation is executed by numerically analyzing the simultaneous differential equations. The simultaneous differential equations can be analyzed using, for example, the Runge-Gutta method.

  When the PBPK simulation is completed, the simulation apparatus 10 records the simulation result in the data storage unit 21 (S40). As a simulation result, data indicating transition (temporal change) of the compound concentration in each human tissue is obtained.

2.3 Verification results The results of verifying the results obtained by the PBPK simulation described above are shown below.

2.3.1 Results from PK Prediction Based on New PBPK Theory Each graph shown in FIG. 9 shows a simulation result of time course of plasma concentration (Ca) in vein for various compounds. In each graph, the horizontal axis represents time, and the vertical axis represents plasma concentration. Further, the “◯” symbol is measured value data, and the solid line indicates the simulation result.

  Referring to the graphs of FIGS. 9A to 9L, it can be seen that good simulation results are obtained for the actual measurement data. For example, very accurate simulation results are obtained for Haloperidol (FIG. 9C), Tamsulosin (FIG. 9H), and Bisoprolol (FIG. 9K).

2.3.2 Comparison with the conventional method Each graph shown in FIG. 10 and FIG. 11 shows the comparison results between the prediction method based on the new PBPK theory (hereinafter referred to as “new PBPK method”) and some conventional methods. It is the figure shown by contrast.

  FIG. 10 is a diagram showing the divergence (fold change) from measured values for several parameters calculated based on PK parameters (plasma concentrations) calculated using the new PBPK method for a plurality of compounds. is there. 10 (a) to 10 (f) also show the degree of divergence of the parameters calculated based on the PK parameters (such as plasma concentration) calculated using the conventional PBPK method for comparison. 10 (a) to 10 (f), a broken line extending parallel to the horizontal axis is a line indicating a boundary (upper limit and lower limit) within a range in which the degree of divergence is within three times, and a one-dot chain line is in which the degree of divergence is within twice. It is a line which shows the boundary (upper limit and lower limit) of a range (same also in FIG. 11). FIG. 10A shows the area under the blood drug concentration-time curve (AUCinf) calculated by extrapolating from time 0 to infinite time. FIG. 10 (b) is for the area under the blood drug concentration-time curve (AUCall) calculated from time 0 to the final blood collection time. FIG.10 (c) is the figure which showed the deviation degree with respect to whole body clearance (CLtot). FIG. 10 (d) is a graph showing the degree of deviation from the plasma half-life (t1 / 2). FIG. 10 (e) is a graph showing the degree of deviation from the distribution volume (vdss) of the compound in a steady state. FIG. 10 (f) is a graph showing the degree of divergence with respect to the plasma compound concentration (Clast) at the final blood collection time.

  10 (a) to 10 (f), the value calculated based on the PK parameter calculated by the new PBPK method according to the present embodiment is the actually measured value than the value calculated based on the PK parameter obtained by the conventional method. It can be seen that the degree of deviation from is small (accuracy is good).

  FIG. 11 compares the new PBPK method with the Dedrick method (rat, dog and monkey). 11A to 11F show the divergence degrees with respect to AUCinf, CLtot, t1 / 2, and Vdss, respectively. From FIG. 11, the value calculated based on the data obtained by the new PBPK method according to the present embodiment has a smaller degree of deviation from the actually measured value than the value calculated based on the data obtained by the Dedrick method (accuracy Is good).

  As described above, according to the new PBPK method of the present disclosure, it is possible to obtain the time transition of the plasma concentration more accurately than the conventional method, and furthermore, various parameters such as AUC, CL, etc. can be obtained accurately based on that. I was able to.

(This disclosure)
As described above, the above embodiment discloses the following prediction (simulation) method.

(1) A method for predicting (simulating) the transition of a human plasma compound concentration using a computer, and setting predetermined parameters to differential equations related to pharmacokinetics (eg, equations (4.1) to (4.9)) And a step of numerically analyzing a differential equation and calculating a time transition of a human plasma compound concentration.
The predetermined parameters include the following values a), b) and c).
a) Physical property values of the compound,
b) the biological value of the compound in relation to model animals and / or humans,
c) A value calculated from at least one of the in vivo measurement value or the physical property value of the compound, the biological value, and the in vivo measurement value obtained from a pharmacokinetic test using a model animal.

  With the above prediction method, it is possible to predict the transition of the concentration in human PK, specifically, human plasma, with high prediction accuracy.

  (2) The method of (1) includes a step of calculating each tissue transferability value (Kp) of the compound related to the model animal from the physical property value of the compound and the biological value of the compound related to the model animal. (1.1)), a step of calculating the steady state distribution volume value (Vdss) of the compound in the model animal from each tissue migration value (Kp) (Equation (1.2)), and the calculated steady state distribution volume value ( Vdss) and the ratio of the compound's steady-state volume of distribution (Vdss) obtained from a pharmacokinetic study using a model animal to correct each tissue migration value (Kp) of the compound for the model animal A step and (formula (1.3)) may be further included.

  By this method, it is possible to obtain a predictive value with high accuracy for the tissue transferability (Kp) to the model animal.

  (3) In the method of (1), the model animal plasma protein non-binding rate (fu, b, rat) is divided by each tissue migration value (Kp, rat) of the compound related to the model animal. A step of calculating the protein non-binding rate (fu, t, rat) in animal tissue (formula (1.6)), an equation (formula (1.7)) representing the correlation of drug concentration between human tissue and model animal tissue, A step of calculating each tissue transferability value (Kp) of the compound related to human from the protein non-binding rate (fu, b, human) in human plasma (formula (1.8)) may further be included. The correlation equation (Equation (1.7)) is the ratio of human tissue volume (Vt, human) to human tissue protein non-binding rate (fu, t, human), model animal tissue volume (Vt, rat) It is a formula which shows the relationship between the ratio of protein non-binding rate (fu, t, rat) in model animal tissue.

  By this method, the human tissue transferability value (Kp) can be calculated from the tissue transferability value (Kp, rat) of the model animal.

  (4) In any one of the methods (1) to (3), the differential equation relating to pharmacokinetics may include a differential equation relating to metabolic kinetics (equation (4.4)).

  (5) In the method of (4), the differential equation relating to metabolic kinetics (formula (4.4)) is a formula for calculating the time change of the compound concentration in human liver, and is a protein non-binding compound in the liver May include a term (= velim, formulas (2.8), (2.14)) including a ratio (fu, liver, fu, p-app). The term (velim) containing the ratio of the compound takes into account the metabolic clearance in the liver (CLint) and / or the action of promoting the uptake of the albumin-bound compound into the liver, and the biological properties of the compound related to humans. It may be calculated using a target value.

  By this method, it is possible to predict the disappearance process with high accuracy, and as a result, it is possible to accurately predict the transition of plasma concentration.

  (6) In any method of (1) to (5), the physical property value of the compound is, for example, the partition coefficient (logP) and / or the acid dissociation constant (pKa) of the compound.

  (7) In any of the methods of (1) to (6), the biological value is, for example, model animal plasma protein non-binding rate (fu, p), model animal blood-plasma concentration ratio (Rb), Model animal tissue protein non-binding rate (fu, t), human plasma protein non-binding rate (fu, p), human blood-plasma concentration ratio (Rb), human tissue protein non-binding rate (fu, t) and Metabolic activity by human liver microsomes or human hepatocytes.

  (8) In any one of the methods (1) to (7), the in vivo measurement value is, for example, the steady state distribution volume (Vdss) of the compound in a model animal, or the steady state distribution volume (Vdss) of the compound in a human. , Human tissue volume (Vt), human blood flow rate (Qt), human glomerular filtration rate (GFR), oral absorption of the compound in the model animal (F) and / or absorption rate constant of the compound in the model animal ( Ka).

  (9) In the method of (8), the model animal is, for example, a rat (an animal other than a rat may be used).

  (10) In any one of the methods (1) to (9), the tissue in the tissue distribution may be one or more tissues selected from lung, skin, fat, kidney, muscle, liver, and intestinal tract.

  (11) The above embodiment discloses a simulation apparatus that executes any one of the prediction methods (1) to (10).

  (12) Said embodiment discloses the program for making a computer perform the prediction method in any one of (1)-(10).

  Since the above-described embodiment is for illustrating the technique in the present disclosure, various modifications, replacements, additions, omissions, and the like can be made within the scope of the claims or an equivalent scope thereof.

DESCRIPTION OF SYMBOLS 10 Simulation apparatus 11 Control part 17 Display part 19 Operation part 21 Data storage part 25 Interface

Claims (12)

A method for predicting the transition of a compound concentration in human plasma using a computer,
Setting predetermined parameters in the differential equation for pharmacokinetics;
Numerically analyzing the differential equation and calculating the time course of the human plasma compound concentration,
The predetermined parameters are
The physical properties of the compound,
Biological value of the compound related to the model animal and / or human, in vivo measurement value or physical property value of the compound obtained from a pharmacokinetic test using the model animal, the biological value, and the in vivo measurement value And a value calculated from at least one of the prediction methods. Calculating each tissue migration property value of the compound related to the model animal from the physical property value of the compound and the biological value of the compound related to the model animal;
Calculating the steady state distribution volume value of the compound in the model animal from the calculated tissue migration value;
Using the ratio between the calculated steady-state distribution volume value and the steady-state distribution volume value of the compound obtained from the pharmacokinetic test using the model animal, each tissue migration value of the compound related to the model animal is corrected. The prediction method according to claim 1, further comprising a step. Calculating the protein non-binding rate in the model animal tissue by dividing the protein non-binding rate in the model animal plasma by each tissue migration value of the compound related to the model animal;
Further comprising the step of calculating each tissue migration value of the compound related to humans from a formula representing the correlation of drug concentration between human tissue and model animal tissue and the protein non-binding rate in human plasma,
The expression indicating the correlation is an expression indicating the relationship between the ratio of the human tissue volume and the protein non-binding rate in the human tissue and the ratio of the model animal tissue volume and the protein non-binding rate in the model animal tissue.
The prediction method according to claim 1.   The prediction method according to claim 1, wherein the differential equation related to pharmacokinetics includes a differential equation related to metabolic kinetics. The differential equation relating to metabolic kinetics is an equation for calculating the time change of the compound concentration in human liver, and has a term including the proportion of non-protein-bound compound in the liver,
The term including the ratio of the compound is calculated using the physical property value of the compound and the biological value of the compound related to humans, taking into account the metabolic clearance in the liver and / or the action of promoting the uptake of albumin-binding compound into the liver. To be
The prediction method according to claim 4.   The prediction method according to claim 1, wherein the physical property value of the compound is a partition coefficient and / or an acid dissociation constant of the compound.   Biological values are model animal plasma protein non-binding rate, model animal blood-plasma concentration ratio, model animal tissue protein non-binding rate, human plasma protein non-binding rate, human blood-plasma concentration ratio, human tissue The prediction method according to claim 1, which is a protein non-binding rate and / or metabolic activity by human liver microsomes or human hepatocytes.   In vivo measurements are steady state distribution volume of the compound in model animals, steady state distribution volume of the compound in humans, human tissue volume, human blood flow rate, human glomerular filtration rate, oral absorption of the compound in model animals The prediction method according to any one of claims 1 to 7, which is an absorption rate constant of the compound in a model animal.   The prediction method according to claim 1, wherein the model animal is a rat.   The prediction method according to any one of claims 1 to 9, wherein the tissue in the tissue distribution is one or more tissues selected from lung, skin, fat, kidney, muscle, liver and intestinal tract.   The simulation apparatus which performs the prediction method in any one of Claims 1-10.   The program for making a computer perform the prediction method in any one of Claims 1-10.

Images