- Email: [email protected]
Estimation of the minimum permeability for perfusion-limited distribution in whole-body physiologically based pharmacokinetic models
Abstracts / Drug Metabolism and Pharmacokinetics 32 (2017) S27eS107
P231 ESTIMATION OF THE MINIMUM PERMEABILITY FOR PERFUSIONLIMITED DISTRIBUTION IN WHOLE-BODY PHYSIOLOGICALLY BASED PHARMACOKINETIC MODELS Yoo-Seong Jeong, Chang-Soon Yim, Heon-Min Ryu, Jun-Hyeng Son, YooKyung Song, Wooin Lee, Dae-Duk Kim, Suk-Jae Chung. College of Pharmacy, Seoul National University, Seoul, South Korea Despite its practical utility, physiologically based pharmacokinetics (PBPK) is somewhat inept in describing the rate of drug distribution to tissues: Membrane-limited distribution is much less frequently used in comparison to more popular perfusion-limited distribution, because the membrane model is applied only when the prediction with the other model is inadequate. In particular, the biopharmaceutical requirement for perfusion-limited model is not adequately determined in the literature. Therefore, the objective of this study was to determine permeability condition for selecting the perfusion-limited model. Since there are two separate equations for the description of the tissue distribution rate, we reasoned that consolidation of these two equations was essential for the study of the rate determining step. We applied two consolidation techniques to PBPK to obtain spatially distributed (Model 1) and spatially homogenized model (Model 2). As expected, these consolidated equations were readily simplified to perfusion-limited and membrane-limited distribution, depending on the permeability surface area product (PS). Series of integration plot analyses were carried out for 11 major tissues (viz, liver, brain, kidney, heart, lung, spleen, gut, muscle, adipose, skin and bone) to evaluate PS values of theophylline, a model drug that primarily behaves as membrane-limited model. Since PAMPA permeability can be easily determined and/or predicted, we reasoned that this permeability could be used as a representative permeability. Accordingly, the surface area, corrected for PAMPA permeability [in unit of m2/g tissue, 0.843 (skin) - 37.4 (kidney) for Model 1, 0.902 (skin) - 39.7 (kidney) for Model 2] was estimated for the major tissues. To determine whether the two consolidated models are predictive to other drugs, we compared the kinetics of in-house / literature observations with the theoretical prediction for 24 drugs with varying permeability. The deviation of the prediction from the observation as expressed by the sums of squared differences indicates that the current models are adequately predictable. Assuming that the rate constant of distribution to tissues of 90% of the perfusion rate is a reasonable indication of perfusion-limited distribution in PBPK, our estimation indicates that, for drugs having the free fraction of unity, PAMPA permeability of 17.2 x 10-6 cm/sec or greater may be necessary for perfusion-limited distribution to all major tissues. P232 IMPROVING THE PREDICTION OF ORAL BIOAVAILABILITY USING FRESH HUMAN INTESTINAL TISSUE Paul E. Cizdziel 1, David C. Bunton 2, Misturu Inamura 1, Graeme C. Macluskie 2, Karen C. McDaid 2, Michael C. Finch 2. 1 Reserach and Development, ReproCELL, Inc., Yokohama, Japan; 2 Biopta Company, Glasgow, Scotland, UK Orally administered drugs continue to be the most common route of drug therapy. It is also increasingly recognised that intestinal biology not only influences absorption but is also an important site of metabolism that influences oral bioavailability. Before a drug progresses to the clinic an estimation of the fraction reaching the systemic circulation is required to optimise the first in man dose. At present, these predictions routinely rely on inputs from animal in vivo or cell based assay models. While the implementation and utilisation of models such as Caco-2 have greatly increased prediction performance over the years, the predictions are recognised to be superior for IV profiles over the more relevant PO profiles for the same set of drugs. This suggests a need to better model the intestinal absorption and metabolism of oral drugs in humans. Here we present data obtained using fresh human small intestinal mucosa tissue mounted in Ussing Chambers. Experiments were conducted to assess the permeability of a range of compounds with the aim of assessing the correlation of the permeability coefficient obtained in vitro to the reported clinical human fraction absorbed (Fa) values. The test compounds were chosen to cover as
S93
wide a range of reported Fa values as possible (12 to 100%). All test compounds were tested at a concentration of 10 mM in an environment maintained at 37 C and pH 7.4. Test compound levels were measured via LC-MS/MS methodology. When comparing the data sets, a steep sigmoidal relationship was observed between in-vitro permeability coefficients and reported Fa values (Hill slope ¼ 2). Although steep, there was a good window between the highest (6.66x10-6 cm/s) and lowest (0.39 x10-6 cm/ s) permeability values observed in vitro. Complementary experiments were also conducted to assess the metabolic activity and Pgp efflux transporter activity. Phase 1 metabolic activity was assessed using model cytochrome p450 (CYP) substrates midazolam (CYP3A4), tacrine (CYP1A2), bufuralol (CYP2D6), diclofenac (CYP2CP) and mephenytoin (CYP2C19). Phase 2 metabolic activity was assessed using model substrates coumarin and diclofenac. Hydroxy, glucuronide and sulphate metabolite formation was measured, again via LC-MS/MS. Pgp activity was assessed using digoxin as a model Pgp substrate and its permeability measured in both apical-basolateral and basolateral-apical directions. Both metabolic (phase 1 and phase 2 enzyme systems) and transporter activities were shown to be preserved in these experiments. The data supports and expands upon the findings of other researchers. In summary, Ussing chambers and fresh human gastrointestinal tissue offer the opportunity to model human absorption whilst taking into account physiologically relevant intestinal metabolism and transporter effects. Ussing chambers also allow the opportunity to directly compare and understand regional and/or preclinical species differences in intestinal absorption and metabolism. P233 LC-MS/MS ASSAY OF MONO BUTYL PHTHALATE, MAJOR TOXIC METABOLITE OF DI-BUTYL PHTHALATE IN RATS Hyeon Gwan Choi, Min Gi Kim, Youngsung Lee, Junwoo Park, Sungwook Park, Sun Dong Yoo. Sungkyunkwan University, Suwon, South Korea Di-butyl phthalate (DBP) is a plasticizer used for manufacturing vinyl foams and other polymers including paints, cosmetics, glue and plastic wraps. DBP is an endocrine disruptor and has been shown to exhibit developmental and reproductive toxicities. One of the most common sources of non-occupational human exposure is contaminated food from plastic wrap or general environment. Once DBP is orally administered, it is rapidly hydrolyzed to mono-butyl phthalate (MBuP) and absorbed. The aim of this study was to develop a sensitive and selective liquid chromatography-electrospray ionization tandem mass spectrometry (LC-ESI-MS/ MS) assay for the determination of MBuP in rat plasma, tissues and urine. MBuP was separated on a Kinetex biphenyl column (100 mm 2.1 mm, 2.6 mm) using a binary gradient mobile phase composed acetonitrile and distilled water. The multiple reaction monitoring was based on the transition of m/z ¼ 223.0/149.0 for MBuP and m/z ¼ 317.0/91.0 for internal standard. The assay was linear in the concentration ranges of 5 e 1000 ng/ ml and validated to demonstrate the linearity, precision, accuracy, LLOQ and recovery by using the matrix matched QC samples. The developed assay was applied to characterize the disposition of DBP after oral administration in rats. Findings of this study may be useful to evaluate the relationship between exposure and toxic potential of DBP in risk assessment. P234 METABOLISM AND PHARMACOKINETICS OF VICAGREL, A NOVEL THIENOPYRIDINE P2Y12 INHIBITOR, COMPARED WITH CLOPIDOGREL IN HEALTHY CHINESE SUBJECTS Cai Liu, Xiaoyan Chen, Dafang Zhong. Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai, China The thienopyridine antiplatelet agent clopidogrel is widely used in the treatment of acute coronary syndrome. Clopidogrel is a prodrug, which requires bioactivation to form its pharmacologically active thiol metabolite (H4) in vivo to prevent platelet aggregation through irreversibly inhibiting the platelet receptor P2Y12 [1]. However, clopidogrel has shown significant interindividual variability in its efficacy mainly caused by the insufficient generation of H4 [2]. Vicagrel, (S)-methyl 2-(2-acetoxy6,7-dihydrothieno[3,2-c]pyridin-5(4H)-yl)-2-(2-chlorophenyl)-acetate, a
P231 ESTIMATION OF THE MINIMUM PERMEABILITY FOR PERFUSIONLIMITED DISTRIBUTION IN WHOLE-BODY PHYSIOLOGICALLY BASED PHARMACOKINETIC MODELS Yoo-Seong Jeong, Chang-Soon Yim, Heon-Min Ryu, Jun-Hyeng Son, YooKyung Song, Wooin Lee, Dae-Duk Kim, Suk-Jae Chung. College of Pharmacy, Seoul National University, Seoul, South Korea Despite its practical utility, physiologically based pharmacokinetics (PBPK) is somewhat inept in describing the rate of drug distribution to tissues: Membrane-limited distribution is much less frequently used in comparison to more popular perfusion-limited distribution, because the membrane model is applied only when the prediction with the other model is inadequate. In particular, the biopharmaceutical requirement for perfusion-limited model is not adequately determined in the literature. Therefore, the objective of this study was to determine permeability condition for selecting the perfusion-limited model. Since there are two separate equations for the description of the tissue distribution rate, we reasoned that consolidation of these two equations was essential for the study of the rate determining step. We applied two consolidation techniques to PBPK to obtain spatially distributed (Model 1) and spatially homogenized model (Model 2). As expected, these consolidated equations were readily simplified to perfusion-limited and membrane-limited distribution, depending on the permeability surface area product (PS). Series of integration plot analyses were carried out for 11 major tissues (viz, liver, brain, kidney, heart, lung, spleen, gut, muscle, adipose, skin and bone) to evaluate PS values of theophylline, a model drug that primarily behaves as membrane-limited model. Since PAMPA permeability can be easily determined and/or predicted, we reasoned that this permeability could be used as a representative permeability. Accordingly, the surface area, corrected for PAMPA permeability [in unit of m2/g tissue, 0.843 (skin) - 37.4 (kidney) for Model 1, 0.902 (skin) - 39.7 (kidney) for Model 2] was estimated for the major tissues. To determine whether the two consolidated models are predictive to other drugs, we compared the kinetics of in-house / literature observations with the theoretical prediction for 24 drugs with varying permeability. The deviation of the prediction from the observation as expressed by the sums of squared differences indicates that the current models are adequately predictable. Assuming that the rate constant of distribution to tissues of 90% of the perfusion rate is a reasonable indication of perfusion-limited distribution in PBPK, our estimation indicates that, for drugs having the free fraction of unity, PAMPA permeability of 17.2 x 10-6 cm/sec or greater may be necessary for perfusion-limited distribution to all major tissues. P232 IMPROVING THE PREDICTION OF ORAL BIOAVAILABILITY USING FRESH HUMAN INTESTINAL TISSUE Paul E. Cizdziel 1, David C. Bunton 2, Misturu Inamura 1, Graeme C. Macluskie 2, Karen C. McDaid 2, Michael C. Finch 2. 1 Reserach and Development, ReproCELL, Inc., Yokohama, Japan; 2 Biopta Company, Glasgow, Scotland, UK Orally administered drugs continue to be the most common route of drug therapy. It is also increasingly recognised that intestinal biology not only influences absorption but is also an important site of metabolism that influences oral bioavailability. Before a drug progresses to the clinic an estimation of the fraction reaching the systemic circulation is required to optimise the first in man dose. At present, these predictions routinely rely on inputs from animal in vivo or cell based assay models. While the implementation and utilisation of models such as Caco-2 have greatly increased prediction performance over the years, the predictions are recognised to be superior for IV profiles over the more relevant PO profiles for the same set of drugs. This suggests a need to better model the intestinal absorption and metabolism of oral drugs in humans. Here we present data obtained using fresh human small intestinal mucosa tissue mounted in Ussing Chambers. Experiments were conducted to assess the permeability of a range of compounds with the aim of assessing the correlation of the permeability coefficient obtained in vitro to the reported clinical human fraction absorbed (Fa) values. The test compounds were chosen to cover as
S93
wide a range of reported Fa values as possible (12 to 100%). All test compounds were tested at a concentration of 10 mM in an environment maintained at 37 C and pH 7.4. Test compound levels were measured via LC-MS/MS methodology. When comparing the data sets, a steep sigmoidal relationship was observed between in-vitro permeability coefficients and reported Fa values (Hill slope ¼ 2). Although steep, there was a good window between the highest (6.66x10-6 cm/s) and lowest (0.39 x10-6 cm/ s) permeability values observed in vitro. Complementary experiments were also conducted to assess the metabolic activity and Pgp efflux transporter activity. Phase 1 metabolic activity was assessed using model cytochrome p450 (CYP) substrates midazolam (CYP3A4), tacrine (CYP1A2), bufuralol (CYP2D6), diclofenac (CYP2CP) and mephenytoin (CYP2C19). Phase 2 metabolic activity was assessed using model substrates coumarin and diclofenac. Hydroxy, glucuronide and sulphate metabolite formation was measured, again via LC-MS/MS. Pgp activity was assessed using digoxin as a model Pgp substrate and its permeability measured in both apical-basolateral and basolateral-apical directions. Both metabolic (phase 1 and phase 2 enzyme systems) and transporter activities were shown to be preserved in these experiments. The data supports and expands upon the findings of other researchers. In summary, Ussing chambers and fresh human gastrointestinal tissue offer the opportunity to model human absorption whilst taking into account physiologically relevant intestinal metabolism and transporter effects. Ussing chambers also allow the opportunity to directly compare and understand regional and/or preclinical species differences in intestinal absorption and metabolism. P233 LC-MS/MS ASSAY OF MONO BUTYL PHTHALATE, MAJOR TOXIC METABOLITE OF DI-BUTYL PHTHALATE IN RATS Hyeon Gwan Choi, Min Gi Kim, Youngsung Lee, Junwoo Park, Sungwook Park, Sun Dong Yoo. Sungkyunkwan University, Suwon, South Korea Di-butyl phthalate (DBP) is a plasticizer used for manufacturing vinyl foams and other polymers including paints, cosmetics, glue and plastic wraps. DBP is an endocrine disruptor and has been shown to exhibit developmental and reproductive toxicities. One of the most common sources of non-occupational human exposure is contaminated food from plastic wrap or general environment. Once DBP is orally administered, it is rapidly hydrolyzed to mono-butyl phthalate (MBuP) and absorbed. The aim of this study was to develop a sensitive and selective liquid chromatography-electrospray ionization tandem mass spectrometry (LC-ESI-MS/ MS) assay for the determination of MBuP in rat plasma, tissues and urine. MBuP was separated on a Kinetex biphenyl column (100 mm 2.1 mm, 2.6 mm) using a binary gradient mobile phase composed acetonitrile and distilled water. The multiple reaction monitoring was based on the transition of m/z ¼ 223.0/149.0 for MBuP and m/z ¼ 317.0/91.0 for internal standard. The assay was linear in the concentration ranges of 5 e 1000 ng/ ml and validated to demonstrate the linearity, precision, accuracy, LLOQ and recovery by using the matrix matched QC samples. The developed assay was applied to characterize the disposition of DBP after oral administration in rats. Findings of this study may be useful to evaluate the relationship between exposure and toxic potential of DBP in risk assessment. P234 METABOLISM AND PHARMACOKINETICS OF VICAGREL, A NOVEL THIENOPYRIDINE P2Y12 INHIBITOR, COMPARED WITH CLOPIDOGREL IN HEALTHY CHINESE SUBJECTS Cai Liu, Xiaoyan Chen, Dafang Zhong. Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai, China The thienopyridine antiplatelet agent clopidogrel is widely used in the treatment of acute coronary syndrome. Clopidogrel is a prodrug, which requires bioactivation to form its pharmacologically active thiol metabolite (H4) in vivo to prevent platelet aggregation through irreversibly inhibiting the platelet receptor P2Y12 [1]. However, clopidogrel has shown significant interindividual variability in its efficacy mainly caused by the insufficient generation of H4 [2]. Vicagrel, (S)-methyl 2-(2-acetoxy6,7-dihydrothieno[3,2-c]pyridin-5(4H)-yl)-2-(2-chlorophenyl)-acetate, a