Detail features... 1

Transcription

1 PBPK Modelling and its Applications to Predict Transporter-Mediated Drug-Drug Interactions Masoud Jamei Senior Scientific Advisor, Head of M&S NEDMDG 14 th June 211, New England, USA Different PK models 1 C= C ie -kit Empirical 2 Compartmental Physiological GT Tucker (Basic PK Course) PBPK Modelling Not a New Idea! Teorell 1937 A mathematical expression of processes involved. Understanding of different influencing factors on time and concentration in target tissue(s). Using this knowledge to achieve a desired therapeutic response. Detail features... 1

2 A Whole Body PBPK Lung Adipose Inhaled Dose Venous Blood Bone Brain Heart Kidney Muscle Arterial Blood IV Dose Skin Portal Vein EHC Spleen Gut Dermal Dose PO Dose Dealing with ADME Properties Bioavailability: release, dissolution, stability, permeability, efflux and/or uptake transport, gut wall and hepatic first pass metabolism,... Distribution: unbound fraction, blood flow, efflux and/or uptake transport, organ size, HSA,... Metabolism: unbound fraction, efflux and or uptake transport, enzyme abundace, blood flow, HSA, Heamatocrite, induction, inhibition,... Excretion: renal clearance, re-absorption, ph effects, active transport, biliary secretion, PBPK Models The Need for Systems Pharmacology Approach Systems pharmacology can be defined as the quantitative analysis of the dynamic interactions between several components of a biological system with the goal of understanding the behaviour of the system as a whole, as opposed to the behaviour of its separate components Inherent in the development of systems clinical pharmacology is a need for models that are physiologically realistic enough to provide accurate systems information yet retain sufficient simplicity to be computationally feasible - in other words, what Vicini [9] describes as fit-for purpose models. However, encouraging progress has been made in predicting individual and even population pharmacokinetics from models that combine biological systems information with physicochemical drug characteristics [1]. Clinical Pharmacology & Therapeutics 21 Detail features... 2

3 Systems Approach: Inter-Individual Variability in PK Reference to Success Story! Systems Data Drug Data Trial Design Mechanistic IVIVE & PBPK Population PK (/PD) Covariates & Study Design In Vitro - In Vivo Extrapolation (IVIVE) Mechanistic Models In vitro system in vitro In vitro CLu int Scaling factors Hepatic Clearance Scaling Factor 1 CLu int per g Scaling Factor 2 in vivo CLu int per Scaling Factors in Human IVIVE In vitro CLu int CLu int per g In vitro system Scaling Factor 1 Scaling Factor 2 CLu int per HHEP HLS9 HLM HLC µl.min cells µl.min -1 mg protein X X HPGL S9PPGL MPPGL CPPGL X Weight Detail features... 3

4 The Segmental-Segregated Flow Model CL other Reservoir D R IV DOSE Q int Venous Blood Q int/3 Q int/3 Q int/3 Arterial Blood Qint D s3 CL d2 CL d1 D sb3 D enb3 D s2 CL d2 CL d1 D sb2 D enb2 D s1 CL d2 CL d1 D sb1 D enb1 f QQint/3 (1-fQ)Qint/3 CL d2 CL d1 CL d2 CL d1 CL d2 CL d1 CL int,m3 M en3 D en3 CL int,m2 M en2 D en2 CL int,m1 M en1 D en1 CL GIT3 CL int.sec3 k a3 CL GIT2 CL int.sec2 k a2 CL GIT1 CL int.sec1 k a1 D lum3 D lum2 D lum1 (Tam et al. 23) ORAL DOSE The Advanced Dissolution, Absorption & Metabolism (ADAM) Model Duodenum Jejunum I Jejunum II Ileum I Ileum II Ileum III Ileum IV Colon Luminal Transit Stomach Emptying Segregated Blood Flows Enzymes (CYP3A4) vs Transporters (Pgp) Jamei et al. 29 and Darwich et al. 21 One Segment in the ADAM Model Metabolism Degradation VARIABILITY / DISTRIBUTION: Radius ph / bile Permeability Enzymes Blood flow Transporters GUT WALL DISSOLVED DRUG FINE PARTICLES SOLID DOSAGE Absorption Efflux/Influx Release Dissolution Precipitation Supersaturation Portal Vein LIVER PBPK DISTRIBUTION MODEL Jamei et al. 29 and Darwich et al. 21 Detail features... 4

5 Inter-individual Variability & fa f a vs P eff and T si (R=1.7 cm) 25 12% f a (%) P eff (cm/h) 1 5 T si (h) M Jamei et al, LogP 24, Switzerland Frequency 2 1% 8% 15 6% 1 4% 5 2% % More SI Transit Time (min) Yu et al. (1998) 1 fa (%) Ileum Jejunum Duodenum.1.1 P eff (x 1-4 cm/s) 1 1 Example of Co-Morbid Obesity & Surgery BPD-DS: Bilio-Pancreatic Diversion with Duodenal Switch GBP: Gastric By-Pass (GBP) ADAM with BPD-DS or GBP Stomach Duodenum Jejunum I & II Ileum I Ileum II Ileum III Ileum IV Colon Solid Dosage Fine Particles Dissolved Drug Degradation Enterocytes Pgp etc Metabolism Faeces Portal Vein PBPK Distribution Model Detail features... 5

6 Predicting V ss V ss knowing distribution into individual tissues is (Sawada et al., 1984): V ss V p V e E : P V t t P t:p V p = volume of plasma; V t = tissue (t) volume Erythrocyte : Plasma partition coefficient Tissue : Plasma partition coefficient Ce, E : P C p, C K p Pt : p C ss ss t, ss p, ss Determining Tissue Volumes System data and their inter-correlations Disease Age Gender Weight Height Adipose Erythrocytes Brain Bone Spleen Plasma Gut Heart Kidney Lung Muscle Skin Organs/Tissue Distributions Lungs Stomach GI Tract Heart Detail features... 6

7 Monte Carlo (MC) vs Correlated Monte Carlo (CMC) Sampling A randomly generated subject using MC sampling Randomly generated subjects using CMC sampling The Complexity of Covariate Effects Body Fat Genotypes (Distribution in Population) Ethnicity Disease Renal Function Serum Creatinine Sex (Distribution in Population) Age (Distribution in Population) Heart Volume Volume Body Surface Area Jamei et al. (DMPK, 29) Cardiac Output Height Weight Weight Cardiac Index Brain Volume Intrinsic Clearance MPPGL HPGL S9PPGL CPPGL Enzyme Abundance Prediction of Tissue:Plasma Partition Coefficients (Kp) One of the most challenging aspects of PBPK modelling is prediction of Kp. Two commonly used mechanistic approaches are: 1. Poulin and Theil * (Berezhkovskiy ** correction) 2. Rodgers and Rowland *** Similarities: Passive diffusion and perfusion-limited tissues models Main differences: Separate models according to compound type (pka) Considering ionisation Acidic phospholipid binding (Ka AP ) Extracellular compartment protein binding (Ka PR ) * Poulin and Theil, 22 ** Berezhkovskiy 24 ** Rodgers and Rowland, 25, 26, 27 Detail features... 7

8 Cirrhosis 1 CYPs e.g CYP3A4 Incidence per 1, Demographics 8 % of control Ascites Median age (y) CP-A (6) CP-B (21) CP-C (21)* % of control 4 Johnson et al, Clinical Pharmacokinetics (21) volume Gastric emptying 16 Renal function GFR (ml/min) shift in gastric emptying T 1/2 in cirrhosis Control CP-A(112) CP-B(11) CP-C(12) Albumin (g/l) 2 Albumin 6 CP-A=46 CP-B=53 CP-C= Control CP-A (94)CP-B (133)CP-C (1) (Johnson et al., 21) Modelling Renal Impairment Prediction of sex effect on kinetics of CYP3A4 substrates Prediction of clearance in obesity Prediction of oral bioavailability after bariatric surgery Prediction of PK in pregnancy Clinical Pharmacology & Therapeutics Feb Detail features... 8

9 Clinical Pharmacology & Therapeutics Feb 211 Zhao 211, CPT 25 Summary of PBPK-IVIVE M&S in submissions From July 28 to June 21, the FDA reviewed 7 investigational new drug (IND) and 6 new drug applications (NDA) submissions containing PBPK modeling and simulations conducted by sponsors. In addition, FDA reviewers conducted PBPK modeling and simulations to support clinical pharmacology reviews of another 4 NDA submissions for which the sponsors did not use PBPK. As a comparison, in the 3 years before 28, FDA received only 2 submissions containing PBPK modeling and simulations. Many of the PBPK modeling and simulation evaluations addressed questions relating to DDIs; others addressed pediatric dosing, the impact of hepatic impairment on drug exposure, and the impact of multiple factors on drug exposure. (Zhao et al., 211).66 per year 8.5 per year An Essential Read on PBPK Ann Rev Pharmacol Toxicol 211; 51: Detail features... 9

10 Key Drug Transporters in Key Organs Intestine Lumen MDR1 (P-gp) Enterocyte Blood The Blood-Brain Barrier The Blood-CSF Barrier MDR1 OATP1A2 BCRP Blood MRP4 MRP4 BCRP OATP2B1 (P-gp) luminal basolateral MRP BCRP PepT1, PepT2, OATP1A2, OATP2B1, OCT3, OCTN1, OCTN2, IBAT, CNT1, CNT2, MCT1, MCT4, MCT5 Choroid epithelium Endothelial cells apical abluminal MDR1 (P-gp) Astrocyte feet Brain parenchyma Cerebrospinal fluid (CSF) Kidney Current focus is on these Apical Basolateral (urine) (blood) Kidney cell OATP1A2 OATP1B1 Blood Hepatocyte MDR1 (P-gp) OCTN1 OCTN2 OAT4 URAT1 OAT1 OAT2 OAT3 OATP1B3 OCT1 MRP2 BCRP MATE1 MATE2-K MDR1 (P-gp) MRP2 OCT2 MRP1 MRP3 MRP4 MRP6 MRP3 BCRP Questionnaire Results: Transporters Focus Group (26) What are the main purposes of your assays? DDI Uptake Clearance Saturation Modelling Ranking Relative Expression Factor (REF): Gut REF = 264/114 = 2.4 Troutman & Thakker, Pharm Res 23 Detail features... 1

11 Bias in Estimation of Transporter Kinetic Parameters Correlation between P-gp protein level and K m(app) value of three P-gp substrates, quinidine (filled triangles), verapamil (filled squares) and vinblastine (filled circles). The reanalysis of the experimental data using the new model revealed that the Km values defined for the intracellular concentration were almost the same among the cells expressing various levels of P-gp for each substrate. Shirasaka et al. (28), J Pharm Sci Tachibana et al. (21), Pharm Res See also Kalvass and Pollack (27), Pharm Res and Lumen et al. (21), DMD Bias in Estimation of Transporter Kinetic Parameters The changes in the experimental conditions caused 1 order of magnitude variation in K m,app and a 5-fold difference in V max,app. However, fitting the concentration data into a compartmental model which accounted for the aqueous boundary layers, cell membranes and cellular retention suggested that the P-gp function per se was not altered. It was the differences in the passive transfer of quinidine which changed the apparent transport kinetics. Permeability-Limited Model () Plasma Water P Capillary Blood Interstitial Fluid fu EW P K tp-on K P-off Extracellular Space +ve ph=7.4 +ve K tp-off ph=7.4 K tp-on P +ve Intracellular Space K tiw-in K tiw-out tight junction +ve NP K tnl-on K tnl-off K tap-on K tap-off +ve Bile fu IW Metabolism NL ph=7 AP -ve EW: Extracellular Water IW: Intracellular Water NL: Neutral Lipids NP: Neutral Phospholipids AP: Acidic Phospholipids Detail features... 11

12 Scaling Approach for Transporters : Scaling to and of Hepatocyte Data HEK-293 HepaRG, Oocytes etc. Jmax/Km CLu int T Scaling Factor 1 HHEP In vitro CLu int, T Scaling Factor 2 CLu int, T per g Scaling Factor 3 Clu int, T per REF HHEP HPGL Weight OATP1B1 OATP1B3 Expression per 1 6 Hepatocyte cells REF = Expression per 1 6 cells in vitro system Hepatocyt e OATP1B1 1/2 = Application of Kinetic Models in In Vitro Assays Different processes have an effect on the distribution of a compound: Passive diffusion, Active uptake or efflux, protein binding Kinetic studies in combination with in vitro assays have highlighted the importance of assessment of these processes. Baker 27; Poirer 28; Paine 28 Decision Tree for P-gp Mediated Interaction Bi-directional transport assay with a probe P-gp substrate in Caco-2 or MDR1-overexpressed polarized epithelial cell lines Net flux ratio of the probe substrate decreases with increased concentrations of the investigational drug Net flux ratio of the probe substrate is not affected with increased concentrations of the investigational drug Likely a P-gp inhibitor Poor or non-inhibitor Determine Ki or IC 5 of the inhibitor [I] 1/IC 5 (or Ki) >.1 or [I] 2/IC 5 (orki) > 1 [I] 1/IC 5 (or Ki) <.1 and [I] 2/IC 5 (or Ki) < 1 An in vivo drug interaction study with a P-gp substrate such as digoxin is recommended. An in vivo drug interaction study with a P-gp substrate may not be needed Zhang 28, Membrane transporters in drug development, Nature Reviews, 21 Detail features... 12

13 Inhibitor Concentrations - Verapamil Dose = 12 mg, MW = 454.6, fu=.1, multiple dose, in a 5x5 simulation: [I] 1 = 1.1 μm [I] 2 = 12mg/25mL = 156 μm Systemic Concentration (µm) Mean Values of Systemic concentration in Mean plasma Values of of Enterocyte Concentration Mean of SimValues - of Luminal Concentration of Sim - Sim-Verapamil over Time Verapamil in Jejunum I over Time Verapamil in Jejunum I over Time Enterocyte Concentration of Drug (µm) Time (h) Luminal Concentration of Drug (µm) Time (h) 1,6. 1, Time (h) Bias in Estimation of Transporter Kinetic Parameters Effect of not considering aqueous boundary layer resistance on K t estimates. Effect of not considering aqueous boundary layer resistance on K i estimates. Results indicate that aqueous boundary layer resistance can bias K t and K i estimates from over-expression systems, where the extent of bias is determined by transporter expression level and substrate affinity. Balakrishnan et al. (27), J Pharmacol Exp Ther Transporters Inhibition % inhibition transport of a substrate IC 5 K i CL int,t inh Jmax [I] u Km 1 Cu K i Keogh and Kunta, 26 Inhibitor conc (μm) Eberl et al., 27 IC 5 determination for clarithromycin varied more than 2-fold: Keogh and Kunta, 26: > 1 μm Eberl et al., 27: 4.1 μm Detail features... 13

14 OATP1B1 Inhibition for a Range of Passive Permeability Values 1.6 AUC Ratio CLpd (ml/min/million hepatocytes) Jmax (pmol/min/million cells) = 5 Km (um) = 1 Ki (um) =.1 Both Substrate and Inhibitor Affecting Each Other Substrate Inhibitor Uptake Transporter CYPs/ Efflux Transporters Effect of Multiple Inhibitors on Substrate and Vice Versa Inhibitor 1 Oral Dose 1- fa Gut Lumen Substrate Inhibitor 2 Oral Dose 1- fa Gut Lumen Faeces fa Faeces fa 1- F G Gut Wall 1- F G Gut Wall Portal Vein Portal Vein IPV IPV FG Q PV Q PV FG QPV QPV QHA QHA Iliver 1- FH CLH FH QPV +QHA Systemic Compartment Isys CLR 1- fa Faeces Gut Lumen fa Gut Wall Portal Vein I PV Oral Dose I liver 1- F H CL H FH Q PV +Q HA Systemic Compartment Isys CL R Metabolites Drug in Urine 1- F G Metabolites Drug in Urine F G Q PV Q PV QHA Systemic Compartment Iliver F H I sys Metabolites 1- F H CLH Q PV +Q HA Drug in Urine CL R Inhibitor 3 Detail features... 14

15 Brain 4-Compartmental Model Q Ssink Q Csink Spinal CSF Volume Protein ph Q Sin Q Sout Volume Protein ph Cranial CSF Q bulk PSE Brain mass Volume Protein ph Enzyme CL met Qbrain PSC CL Cin CL Cout Brain blood Volume Protein ph PSB CL Bin CL Bout Qbrain CSF: circulation, proteins, ph, volume, Brain: anatomy, physiology, proteins, enzymes, System parameters BBB and BCSFB permeability values Active transport in any of the barriers Drug-related parameters Acknowledgment - The Simcyp Team Detail features... 15