NEDMDG Fall Meeting September 13, 2017

Recent Advances in Drug Transporter and Enzymology Sciences to Enable Drug Design

Larry Tremaine

Predict Efficacious Concentration

Predict PK

Predict DDI Predict Variability

Target Ki

fu,plasma

fu,in vitro

BL/PL

Kp,uu,target

CLint,met

fmet

CYP IC50

UGT IC50

Transporter IC50

ftransport

CLint,u

Predict Dosing Regimen

fu,target

PD Response

CLint,txp

CYP kinact/KI

Predict Safety

Animal CLrenal

The Parameters We Measure and the Predictions to Which They Contribute

Predict Human

Metabolites

Metabolite ID

TK

Courtesy of S. Obach

ADME Properties and Assays in Drug Design

ADME Property Measurements/Parameters Assays

CL - metabolic Hepatic metabolism, reaction phenotyping

HLM, suspension HHEPs or HHEP relay +/- chemical inhibitors

DDI - victim Fm, Ft Chemical inhibitors with HHEPs, HLM

CL - renal Ppb, active secretion Rat SSS, OATs-HEK RAF

CL – hepatic uptake Active uptake, passive uptake, transporter phenotyping

SCHH, PHH +/- inhibitors, OATPs-HEK RAF; NHP SSS

Ceff, tissue distribution

Ppb; Cbu/Cpu Tissue exposure; fu,cell

In silico; equilibrium dialysis; Pgp and BCRP-MDCK; Kpuu;

CL – biliary Biliary excretion SCHH +/- Ca2+

DDI - perpetrator Competitive DDI TDI Induction

HLM HLM SCHH, PXR reporter assay

Absorption - fafg Permeability; Efflux transporters; CYP3A metabolism

RRCK (low canine Pgp expression) Pgp and BCRP MDCK HLM, preclinical species PK

Genesis of Predicting Human CL from In Vitro Data at Pfizer

PK Predictions – A Long History At Pfizer

PDM CONSISTENTLY re-evaluated our PK-predictions to improve and evolve our Science

There has been many teams over the years

Major Challenge Projected clinical PK

parameters with LIMITED Human IV

Data

Courtesy S. Steyn Obach et al.,JPET (1997) 283:46-58 Hosea et al., (2009) J Clin Pharmacol 49:513-533

Changing Philosophy to Predicting Human PK

0

20

40

60

80

100

120

1 10 100

% o

f com

poun

ds

Fold error

rat sss

dog sss

NHP SSS

HLM PBPK Vss

HHEPS PBPK Vss

No single method accurately predicts ALL compounds (2006-2010 Data Set) If several approaches agreed, then high “confidence in prediction”

Extended Clearance Classification System Predictability

7

Bases/NeutralsAcids/Zwits

Hig

h P

erm

eabl

eLo

w P

erm

eabl

e

Class 1

Class 3

Class 2

Class 4

Metabolism

Renal

Class 1A Class 1BMetabolism Hepatic uptake

MW ≤400 MW >400 n=172

n=32

n=29 n=14

Renal5%

Metabolism95%

Renal75%

Metabolism25%

Renal10%

Metabolism90%

Renal14%

Hepatic uptake

86%

Class 3A Class 3BRenal Hepatic uptake

(or) Renal

MW ≤400 MW >400

n=24n=36

89%

Obtained data from Obach et al. + other papers.

N = 789

Identified rate-determining clearance step using Pfizer/Novartis dataset

N = 313

Utilized phys chem/ in vitro properties to classify compounds based on their rate determining clearance step

Varma MV, Steyn SJ, Allerton C, El-Kattan AF. Predicting Clearance Mechanism in Drug Discovery: Extended Clearance Classification System (ECCS). Pharm Res. 2015, 3785-3802

Permeability cut off = 5 X 10-6 cm/sec Molecular Weight cut off:

400 [Class 1] Ionization:

Acid+Zwit vs Base+Neutral

Prevalence and importance of non-P450 metabolism

2 0 0 52 0 0 6

2 0 0 72 0 0 8

2 0 0 92 0 1 0

2 0 1 12 0 1 2

2 0 1 32 0 1 4

2 0 1 52 0 1 6

0

5

1 0

1 5

2 0

2 5

Dru

gs

N o M a jo r M e ta b o lite sP 4 5 0N o n -P 4 5 0

1 09

8

1 2

7

1 3

1 7

1 4

1 8 1 7

6 6

M Cerny DMD 2016, 44:1246–1252

FDA–Approved Oral and Intravenous Drugs: 2006–2015

Clearance Panel – Identify major CL mechanism, appropriate in vitro systems and generate rate constants

HHeps+/-ABT

Empirical Classification based on Physiochemical Properties

Transporters Acids

OATPs, OATs

CYP Dominated

Non-CYP Dominated

rCYP CLint UGT CLint

AO CLint

Other Conj.

Other Ox.

Hydro-lysis

Tier 1

Tier 2

No Metabolism

HHeps Type of metabolism

Transporters Bases OCTs

ECCS Class 2 and 1A ECCS Class 1B, 3 and 4

Current Chemistry Space - Metabolic HHEP CLia

From Jan, 2016 – June 2017 (18 months) approximately 7000 compounds submitted to 4 hr suspension HHEP for metabolic CL Approximately 1000 compounds have Clia <3 ul/min/million cells.

~ 50% Low

Clia<12

~ 30% High Clia>20

~ 20% Moderate 12<Clia<20

Courtesy of A. Carlo

Extending Clmetabolic Dynamic Range – Hepatocyte relay method for low clearance

Compounds Enzyme In vivo Relay MethodDiazepam CYP3A, 2C19 15 15

Tolbutamide CYP2C9 4.9 6.9Theophylline CYP1A2 2.1 2.5

Timolol CYP2D6 36- 49 14Disopyramide CYP3A 5.9 4.6

S-Warfarin CYP3A, 2C9 4.5 5.1Zolmitriptan CYP1A2 and MAO 13 3.2*

CLint (mL/min/Kg)

Transfer Supernatant

* Under-predict due to extra-hepatic contributions of MAO

Di et al., DMD (2012) 40:1860-1865 Di et al., DMD (2013) 42:2018-2023

Assays Incubation Time CL int

Detetion Limit

Tier 2a HHEP 4 Hours 7.5 mL/min/Kg

HHEP Relay 20 Hours – 0.5Mcells/ml 20 Hours – 2.0Mcells/ml

1.46 mL/min/Kg 0.36 mL/min/Kg

IVIVE for Metabolic CL using Global Lot of HHEPs

CLb,obs ≤ 10 mL/min/Kg

HHEP Relay

HHEP suspension

Hepatocyte lot DCM apparent intrinsic clearance (𝑪𝑪𝑪𝑪𝐢𝐢𝐢𝐢𝐢𝐢,𝒂𝒂𝒂𝒂𝒂𝒂) measures should be scaled by a factor of 1.0 regardless of assay conditions (4 hr suspension or relay)

IVIVE for CLb,obs ≤ 10 mL/min/Kg • Most applicable clearance range for oral drugs. • 70% of data within this cutoff (need high n’s for high confidence) • Cutoff enables application of single scalar regardless of assay type.

All Data, N = 35 compounds

Courtesy S. Steyn

Metabolic Phenotyping - Reporting Limitations of Substrate Depletion Assays (CYP)

• Highlighted values represent compound CL required to enable capture of majority of CYP activity under given conditions (sum of ALL contributing isoforms)

– i.e. 90% of one enzyme, 90% of 2 enzymes (40% CYP1, 50% CYP2)

• Increased sensitivity comes with increased reagent consumption and assay cost

Current limit of parent

depletion

Increasing sensitivity & resource

Courtesy of A. Doran and S. Obach

Assay HLM HHEP HLM HHEP Relay HHEP Relay Incubation time 60 min 4 H 60 min 20 H 20 H Concentration 0.75 mg/mL 0.5e6 cells/mL 2.0 mg/mL 0.5e6 cells/mL 2e6 cells/mL Lowest Reportable CLi,a,s 7.0 mL/min/kg 7.5 mL/min/kg 2.6 mL/min/kg 1.5 mL/min/kg 0.36 mL/min/kg

Measured CLint,a,s 150 95% 94% 98% 99% 100% 40 83% 90% 94% 96% 99% 30 77% 86% 91% 95% 99% 20 65% 79% 87% 93% 98% 10 30% 58% 74% 85% 96% 9 22% 53% 71% 83% 96% 7 0% 40% 63% 79% 95% 6 -- 30% 57% 75% 94% 5 -- 16% 48% 70% 93% 4 -- 0% 35% 63% 91% 3 -- -- 13% 50% 88% 2 -- -- 25% 82% 1 -- -- 64%

0.8 -- -- 55% 0.6 -- -- 40% 0.4 -- -- 10%

Maximum Observable Inhibition

Standard Screening Assays

Pre-incubation with Inhibitor

Remove Inhibitor Incubate with Substrate

Transfer Supernatant

Next Relay

Novel HHEP Relay - Reaction Phenotyping Assay

MBIs to avoid the impact of inhibitor depletion during long incubation

MBIs were removed from incubation after inactivation to increase selectivity by

minimizing reversible inhibition (CYP, AO, UGT, SULT, CES, Transporters)

Add Substrate

Courtesy of L. Di

Yang X, Atkinson K, Di L., DMD (2016) 44(3):460-5

Experimental Conditions for HHEP Relay Phenotyping

CYP Pre-Incub. Time (min)

Inactivators and Concentrations Substrates Metabolites

1A2 15 1 µM Furafylline Phenacetin Acetaminophen

2B6 15 10 µM Phencyclidine Bupropion OH-Bupropion

2D6 15 1.8 µM Paroxetine Dextromethorphan Dextrorphan

2C8 30 100 µM Gemfibrozil Glucuronide Amodiaquine N-Desethylamodiaquine

2C9 30 15 µM Tienilic Acid Diclofenac 4′-OH-Diclofenac

2C19 15 8 µM Esomeprazole Mephenytoin 4′-OH-Mephenytoin

3A 15 25 µM Troleandomycin Midazolam 1′-OH-Midazolam

3A4 15 2 µM CYP3cide* Midazolam 1′-OH-Midazolam Pan-CYP 30 1 mM ABT & 15 µM

Tienilic Acid All Above All Above

* Genotyped HHEP CYP3A5 EM

Courtesy of L. Di

Yang X, Atkinson K, Di L., DMD (2016) 44(3):460-5

HHEP Relay Phenotyping Assay Validation

Reported fm : 0.5 3A, 0.5 2C19

Relay fm: 0.56 3A, 0.42 2C19

1

10

100

1000

0 5 10 15 20 25

% R

em

ain

ing

Time (hour)

Diazepam (3A + 2C19)

1

10

100

1000

0 5 10 15 20 25

% R

em

ain

ing

Time (hour)

Diazepam + TAO

1

10

100

1000

0 5 10 15 20 25

% R

em

ain

ing

Time (hour)

Diazepam + Esomeprazole

Courtesy of L. Di

Yang X, Atkinson K, Di L., DMD (2016) 44(3):460-5

Assay Radiometric Mass SpecIncubation time 30 min 30 minConcentration 2 mg/mL pt HLM 2 mg/mL pt HLMSubstrate Conc 1 uM 1 uMQuantititive limit 200 - 100000 dpm (0.2%) 0.1 nM (LLOQ) Lowest Reportable CLi,a,s 0.03 mL/min/kg 0.0015 mL/min/kg

Measured CLint,a,s4 100% 100%

0.4 99% 100%0.04 93% 96%0.01 70% 85%

0.004 25% 63%0.002 25%

Maximum Observable Inhibition

Metabolite Formation

Increased Sensitivity of Metabolite Formation Assays (CYP)

• Affords increased dynamic range due to differences in bioanalytical endpoint (change in metabolite response), decreased reagent usage.

• Availability of radiolabel compound offers additional opportunities for assessing low CL phenotyping.

Parent depletion limit

Increasing sensitivity

Courtesy of A. Doran and S. Obach

An Overview of the Isolation Process Walker, et al, DMD, 2014 42:1627-1639

Metabolite Biosynthesis and NMR quantitation

• Transporters are important determinants of clearance (DDI) & exposure at the site of efficacy / toxicity

• Common target tissues: Brain, Liver, Kidney & Intestine

Opening design space Transporter space as a key gap and opportunity

Courtesy of M. Varma

Varma, MV et al. Pharm Res (2015), 32(12), 3785-3802. El-Kattan, AF et al. Pharm Res (2016), 33(12), 3021-3030 Varma, MV. et al. Adv Drug Deliv Rev (2017), epub Ahead of Print.

~ 20 % of design effort in complex disposition space High potential for exposure asymmetry (opportunity and/or liability)

Active Renal Excretion A relevant space with room for improvement

Courtesy of M. Varma

Predicting CLrenal via Single Species Scaling

• No mechanistic description of renal reabsorption or active renal secretion

from Rat CLr from Dog CLr

Correction for plasma protein binding Correction for plasma protein binding and kidney blood flow

Paine et al., DMD 2011, 1008

Courtesy of M. Varma

OAT1 Substrate selectivity

------------ OAT2 Substrate selectivity --------------------

--------------------------- OAT3 Substrate selectivity ----------------------------------

0 2 40

5

1 0

1 5

2 0

T e n o fo v ir

T im e(m in )

To

tal

Ce

ll A

mo

un

t(p

mo

l)

W T

h O A T 1

h O A T 3

0 2 40

2

4

A c y c lo v ir

T im e(m in )

To

tal

up

tak

e(p

mo

l/mg

of

pro

tein

)

W T

h O A T 1

h O A T 3

h O A T 2

0 2 40

2

4

6

G a n c ic lo v ir

T im e(m in )

To

tal

up

tak

e(p

mo

l/mg

of

pro

tein

)

W T

h O A T 1

h O A T 3

h O A T 2

0 2 40

1

2

3

O s e lta m iv ir a c id

T im e(m in )

Ce

ll u

pta

ke

(pm

ol/m

g p

rote

in)

W T

h O A T 1

h O A T 3

0 .0 0 .5 1 .0 1 .5 2 .00

2 0

4 0

6 0

8 0

1 0 0

F u r o s e m id e

T im e(m in )

Ce

ll u

pta

ke

(pm

ol/m

g p

rote

in)

W T

h O A T 1

h O A T 3

0 2 40

2

4

6

8

1 0

B e n z y lp e n ic i ll in

T im e(m in )

Ce

ll u

pta

ke

(pm

ol/m

g p

rote

in)

W T

h O A T 1

h O A T 3

Time-course of uptake by OATs-transfected HEK293 and wild-type cells

Mathialagan et al., DMD 2017, 409-417 RAFOAT1=0.64

RAFOAT2=7.3 RAFOAT3=4.1

CLrenal - OAT scaling via single transfect cells

Mathialagan et al., DMD 2017, 409-417

• Descriptions of both renal reabsorption and active renal secretion

Courtesy of M. Varma

Active Renal Excretion Towards a fit for purpose approach to clearance prediction

Courtesy of M. Varma

Mathialagan et al., DMD 2017, 409-417

Hepatobiliary Transport A relevant space with room for improvement

Giacomini KM & Huang S-M (2013) CPT Transporters in Drug Development and Clinical Pharmacology

Varma, MV et al. Pharm Res (2015), 32(12), 3785-3802. El-Kattan, AF et al. Pharm Res (2016), 33(12), 3021-3030 Varma, MV. et al. Adv Drug Deliv Rev (2017), epub Ahead of Print.

Sirianni GL & Pang (1997) J Pharmacokin Biopharm 25(4) 449 - 470

~ 15% of design effort in hepatobiliary transport disposition space High potential for exposure asymmetry (opportunity and/or liability)

Courtesy of M. Varma

In vitro tool – Primary human hepatocytes

26

Suspension Heps (Oil spin method)

Plated human hepatocytes (PHH)

Sandwich-cultured human hepatocytes (SCHH)

Before 2004

Circa 2005

2015-16

UNC/Pfizer collaboration Bi et al. DMD 2006 2011/12

SCHH-PBPK model Jones et al. DMD 2012

Adopted from Sugiyama et al.

2014 Middle-out PBPK Li et al. DMD 2012

2012-14 SCHH-DDI M&S Varma et al. 2012

Milestones

Courtesy of M. Varma

Mechanistic PBPK Modeling for Prediction of Hepatic Transporter-Mediated Disposition Using SCHH Data

Jones, et al. DMD (2012) 40-1007-17

Hepatic Transporter Disposition -Current best practice (SCHH/PBPK)

v3 – Li (2014), J Pharmacokinet Pharmacodyn, 41(3), 197-209 – Trained with six ECCS 1 compounds devoid of biliary clearance

• Use SCHH inputs with the following training set. – Bosentan, Cerivastatin, Fluvastatin, Glyburide, PF-05186462 (NAV1.7), &

Repaglinide

– Population fit of clearance scaling factors (𝑆𝑆𝑆𝑆) • Log-normal inter compound variability on scaling factors (𝑆𝑆𝑆𝑆)

– Account for imprecision in compound inputs – Provides an method to determine uncertainty in prospective predictions by

bootstrapping the scaling factor distribution space

0.1

1

10

0.1 1 10

Pred

ictio

ns

Observations

Model SFpass SFact

SFmet

v3 (populati

on)

0.35 27 0.19

90% CI 0.064-2.1

14-49

0.065-0.56

CV 110% 39% 66%

Transporters in vitro HHEP systems do not represent in vivo physiological activity levels

Courtesy of T. Maurer

Hepatic Transporter Phenotyping IC50 OF SELECTIVE INHIBITORS in HEK293 CELL LINES

Yi-An Bi ISSX 2017 poster

Challenging Class 1B compds: Montelukast case

Con tro l (

3 7 °C )

4 °C

+ R ifam

p icin

(3 µ

M)

+R ifam

p icin

(10 µ

M)

+ C y c lo sp

o r in e A

(10 µ

M)

+R ifam

y c in S

V (2

0 µM

)

+R ifam

y c in S

V (1

mM

)

+ Gem

fibro

z il (

1 0 0 µM

)

+Gem

-glu

(10 0 µ

M)

+MK 5 7 1 (5

0 µM

)

0

5 0

1 0 0

1 5 0

2 0 0

% o

f c

on

tr

ol

Up

ta

ke

CL

int

(0

.5-5

m)

ECCS 1B 1

10

100

1000

0 3 6 9 12 15 18 21 24

Mo

nte

luka

st p

lasm

a c

on

c. (

ng

/mL)

Time (h)

1

10

100

1000

10000

0 3 6 9 12 15 18 21 24

Mo

nte

luka

st p

lasm

a c

on

c. (

ng

/mL)

Time (h)

0

0.5

1

1.5

2

Control + Rifampicin

Mo

nte

luka

st C

L (m

L/m

in/k

g)

Interaction with rifampicin in Cynomolgus monkeys

(D) (

Interaction with rifampicin in rats

(A) (

0

5

10

15

20

25

30

Control + Rifampicin

Mo

nte

luka

st C

L (m

L/m

in/k

g)

*P=0.011

*P=0.012

CL reduced with Rifampicin in rat and monkey

Varma et al., CPT (2017) 101:406-415

Montelukast – recommended as substrate for CYP2C8 clinical DDI studies

PK prediction strategy

LUNG

BRAIN

KIDNEY

ADIPOSE

HEART

MUSCLE

OTHERS

SKIN

VEN

OU

S B

LOO

D

ARTE

RIA

L B

LOO

D

LUNG

BRAIN

KIDNEY

ADIPOSE

HEART

MUSCLE

OTHERS

SKIN

VEN

OU

S B

LOO

D

ARTE

RIA

L B

LOO

D

GUT

• CLbile from SCHH • Other ADME inputs from

Enzymology and HDO

Rat/Monkey DDI with Rif

Cyno SS scaling

No

Assume no active uptake in human for

PK predictions

Yes

ECCS

Uptake inhibitable by Rif SV

Class 1B Class 3B

Tier 1 PHH assay

Yes

Phenotyping Characterize transport mechanisms using HEK panel, selective inhibitors, etc.

No

0.1µM conc.*

Renal CL

Class 3B

Class 1B

PBPK modeling

Multiple replicates for dose projections

* Compounds with analytical sensitivity issues will be run at 1µM conc. as Tier 2.

Courtesy of M. Varma

OATP-HEK Uptake and steady state unbound tissue-to-plasma exposure of Glucokinase Activators PF-049 and PF-051 determined in rats post–4-day infusion.

A Ghosh et al. DMD 2014;42:1599-1610

Copyright © 2014 by The American Society for Pharmacology and Experimental Therapeutics

Time (sec)

PF-049 PF-051

Glucokinase Activators

Drug Disposition and Trafficking in Cells – Factors Contributing to Asymmetric Distribution

Mitochondria pH 8

-167 mV

Influx

Passive Diffusion

Lysosomes pH 4.7

+ 19 mV

Cytosol: pH 7.1, -38 mV

D

M1 M2 …

Proteins

Metabolism

Binding

Lipids

pH Gradient

Membrane Potential

Medium/Blood pH 7.4

63%

28%

0.8%

Nucleus pH 7.2

-9.2 mV 8%

Courtesy of L Di

Introduction of Kpuu

Pharmacology and ADMET

Cell-Based

In Vitro

Kpuu = 𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅 𝐂𝐂𝐅𝐅𝐂𝐂𝐂𝐂𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅 𝐌𝐌𝐅𝐅𝐌𝐌𝐢𝐢𝐌𝐌𝐌𝐌

Steady State

Blood Liver

In Vivo

Kpuu = 𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅 𝐓𝐓𝐢𝐢𝐓𝐓𝐓𝐓𝐌𝐌𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅 𝐁𝐁𝐂𝐂𝐁𝐁𝐁𝐁𝐌𝐌

Tissue exposure, PD, TI & DDI

CLint = CLint′

Kpuu EC50 = EC50

’ × Kpuu IC50 = IC50’× Kpuu

Courtesy of L Di

Predict In Vivo Hepatic Kpuu (IVIVE) in Rat

Compounds

In Vivo Rat Liver-

to-Plasma Kpuu

(IV Infusion)

In Vitro

Suspension Rat

Hepatocyte Kpuu

Fold Difference

in Vivo Kpuu /in

Vitro Kpuu

Cerivastatin 29 ± 8.5 21 ± 2.0 1.4

Fluvastatin 44 ± 8.7 21 ± 1.5 2.1

Rosuvastatin 57 ± 9.5 35 ± 0.6 1.6

Pravastatin 2.2 ± 1.5 2.9 ± 0.3 0.8

PF-04991532 5.7 ± 1.6 7.1 ± 0.1 0.8

PF-05187965 2.4 ± 0.6 4.2 ± 0.2 0.6

Good IVIVE for Substrates of Oatps in Rat K Riccardi et al., DMD, (2017), 45:576-580

Approaches to Predicting Human Hepatic CL

Drug Distribution to Brain - physiology of the central barriers

Courtesy of P. Trapa

Title

1 1 0 1 0 0 1 0 0 0

0 .0 1

0 .1

1

1 0

M D R 1 (B o rs t) E ff lu x R a tio

Ag

g.

rod

en

t A

UC

b,u

/p,u

E R = 2 .5

A U C b ,u /p ,u = 0 .3

1 1 0 1 0 0 1 0 0 0

0 .0 1

0 .1

1

1 0

M D R 1 (N IH ) E ff lu x R a tio

Ag

g. r

od

ent

AU

Cb

,u/p

,u

E R = 6 .0

A U C b ,u /p ,u = 0 .3

Borst and NIH MDR1-MDCK Efflux Ratio Correlation with Rodent Brain Penetration (Cbu/Cpu)

Courtesy of B Feng

Title

1 1 0 1 0 0

1

1 0

1 0 0

M D R 1 (B o rs t) E ff lu x R a tio

MD

R1

(N

IH)

Eff

lux

Ra

tio

E R = 2 .5

E R = 6 .0

Borst MDR1 Assay vs. NIH MDR1 Assay

The two MDR1 assays do not correlate Courtesy of B Feng

Title

MDR1 mRNA P-gp Protein

N IHB o rs

tN IH

B o rst

N IHB o rs

t0

5

1 0

1 5

2 0

Rat

io n

orm

aliz

ed b

y B

ors

t

TEER

Key Properties for Borst vs. NIH MDR1 Cells

Courtesy of B Feng

Translational Strategy for Brain Penetration

Exptl. Parameters – Compound Specific

binding: in vitro equilibrium dialysis

intrinsic passive brain permeability: in silico

active transport at the brain/choroid plexus: analysis of in vitro cell lines (MDR1 and BCRP)

compound specific

Output

extracellular fluid (ECF)

intracellular fluid (ICF)

CSF

plas

ma

BBB

BCSFB

species specific

expression levels: measured from tissues and cell lines or activity from IVIVC at steady state

physiology:

from literature

bulk flow active transport passive diffusion

Trapa PE et al. (2016) J Pharm Sci 105: 965 - 971

Key assumptions Steady-state

Passive flux >> bulk ESF flow Only PgP (ER1) & BCRP (ER2)

2 approaches • Model based estimation against neuroPK data • Relative protein expression

Brain Penetration Exhaustive translational model qualification

Courtesy of P Trapa

Efflux Transporter expression-level changes across species can impact distribution

Trapa PE et al. (2016) J Pharm Sci 105: 965 - 971

• Mechanistic Approach to Clearance IVIVE – CLhepatic mature but challenged by expanded chemical space – Incorporation of hepatic- and renal-transporter mediated CL

• Phenotyping – Lower and non-CYP metabolic CL require novel approaches to

phenotyping – Drug transporter phenotyping is still evolving and lacks clinical markers

• Drug Distribution – Greater appreciation of intracellular pH, membrane potential and

uptake and efflux transporters influencing tissue Kpuu and BBB permeation

Summary

• IVIVE relationships are limited by lack of human PK – Cliv; Vd; non-CYP PK; fafg; Cbu/Cpu – Consideration of preclinical species

• Reagent / technology advances continually needed – HHEP relay costs impacts capacity

• Co-cultures, immortalized cell lines and microphysiological systems

– Hepatic transporter cell model(s) needed with greater dynamic range • Can we identify an Industry-wide approach?

• Non-traditional modalities will continue to increase – E.G. – Antibody drug conjugates, biologics, gene therapy, parenteral

delivery

Gaps and Horizon

Acknowledgements Enzymology (and Preclinical PK) Transporter Sciences

Group PDM Project Reps / M&S and Biotransformation

Atkinson, Karen Lin, Jian Bi, Yi-An El-Kattan, Ayman

Cerny, Matt Lin, Zhiwu Costales, Chester Kalgutkar, Amit

Cianfrogna, Julie Lovering, Elaine Tseng Eatmadpour, Soraya Li, Rui

Dantonio, Aly Niosi, Mark Feng, Bo Maurer, Tristan

Di, Li Novak, Jon Kimoto, Emi Scott, Dennis

Eng, Heather Nulick, Kelly Lazzaro, Sarah Steyn, Stefan

Fahmi, Odette Obach, R Scott Mathialagan, Sumathy Tess, David

Doran, Angela Orozco, Christine Rodrigues, Dave Trapa, Patrick

Goosen, Theunis Riccardi, Keith Tylaska, Laurie Walker, Greg

Johnson, Nat Ryu, Sangwoo Varma, Mathena

Kosa, Rachel Strelevitz, Tim Vildhede, Anna

Lapham, Kim Yang, Xin West, Mark

Additional Thanks to Colleagues in Hit Design Lead Optimization, NonReg Bioanalytical and Comparative Medicine

Thank You for Your Attention

Approaches to Predicting Human Hepatic CL