in vitro in vivo extrapolation and its applications in predicting pk … · 2020-06-29 · james r....

In Vitro In Vivo Extrapolation and its

Applications in Predicting PK Population Variability

Alice Ke, PhDConsultant & Scientific Advisor

Simcyp Limited

© Copyright 2013 Certara, L.P. All rights reserved.

Outline

• Clearance concept

• In Vitro In Vivo Extrapolation (IVIVE)

• Linking PBPK and IVIVE, accounting for variability

• Transporters

• Industry/Regulators views

• Future prospects


Well-stirred liver model

James R. Gillette, Ann N Y Acad Sci. 1971

Commentary: A physiological approach to hepatic clearance Wilkinson and Shand , CPT, 1975

Pang and Rowland, JPK Biopharm 1977

fB,Out = Unbound drug in venous blood /Whole emergent blood concentration

• Unbound concentration of drug in blood cells equates to the unbound concentration in plasma.

• Emergent venous blood is in equilibrium with that in the liver.Rowland, Benet and Graham, JPK Biopharm 1973 Yang et al., DMD, 2007


In Vitro - In Vivo Extrapolation (IVIVE)

in vitro CLuint

in vivo CLuint

Scaling factors

Mechanistic Models

CLuint per Liver

CLuint per g Liver

In vitrosystem

In vitroCLuint

Scaling Factor

(MPGGL, HPGL)

Liver weight


Accuracy of IVIVE approaches for human CL or CLint

Generally many literature studies shows under-prediction from in vitro systems.Can be corrected using an empirical scaling factor.Need to understand for your in vitro system if this is necessary.

System AFE RefHLM 2.3 Obach DMD 27, 1350, 1999HLM 6.2 Ito, Pharm Res, 22, 103, 2005HLM 2.3 Stringer, Xeno, 38, 1313, 2008HLM 2.2 Ring, J Pharm Sci, 100, 490, 2011HLM 5 Hallifax, Pharm Res, 27, 2150, 2010HLM 2 Jones, Clin Pk, 50, 311, 2011Heps 2.4 De Buck DMD, 35, 1766, 2007Heps 5.2 Stringer, Xeno, 38, 1313, 2008Heps 5 Hallifax 2011Heps 7.6 Naritomi DMD 31, 580, 2003HepsRecombinant CYP 1.53 PT

2.15 WSStringer DMD, 37,1025, 2009


IVIVE predictions – Improvements over years

• Non-specific binding (Obach, DMD, 1999, Riley et al., DMD, 2005)

• Recombinant CYPs and ISEF values (Galetin et al., DMD, 2004; Proctor

et al. Xenobiotica, 2004)

• In vitro modelling to account for hepatic uptake (Soars et al., DMD, 2007)

• Adding BSA and HAS-FAS to HLM (Rowland et al., DMD, 2008)

• Accounting for the difference in drug ionization in extracellular and

intracellular tissue water (Berezhkovskiy, J Pharm Sci, 2011)

• Integrating uptake, metabolism, biliary excretion, and sinusoidal efflux

(Umehara and Camenisch, Pharm Res, 2012)

• Incorporating ionisation and protein binding (Poulin et al., J Pharm Sci,

2012)


Gut wall metabolism

‘Qgut’ , a minimal model

gutintgutgut

gutg CLu.fu'Q'

'Q'F

−+=

villiperm

villipermgut QCL

QCL'Q'

+

⋅=

Yang et al., CDM, 2007

0

0.2

0.4

0.6

0.8

1

0 0.2 0.4 0.6 0.8 1Observed Fg

Pred

icte

d F g

Gertz et al., DMD, 2010


Special populations

AgeWeight

Tissue VolumesTissue Composition

Cardiac OutputTissue Blood Flows

[Plasma Protein]

Systems Data

Drug Data

TrialDesign

MWLogPpKa

Protein bindingBP ratio

In vitro MetabolismPermeability

Solubility

Dose Administration route

FrequencyCo-administered drugs

Populations

Prediction of drug PK (PD) in population of interest

Mechanistic IVIVE linked PBPK models

Jamei et al., DMPK, 2009, Rostami-Hodjegan, CPT, 2012


Demographic Features of Healthy and Disease Populations

Freq

uenc

y

Age

Randomly Generated

HVDisease

Defined by real data


Age Distribution in Target Population

0

1500300045006000 Addicts

0

200

400

600

800 CVD

MALEFEMALE

Freq

uenc

y

Age Category


The Complexity of Covariate Effects as Applied to CL

Age(Distribution in Population)

Ethnicity Disease

Height

CardiacIndex

MPPGLHPGL

Plasma Proteins

&Haematocrit

SerumCreatinine

Body Surface

AreaWeight

LiverVolume

CardiacOutput

IntrinsicClearance

Renal Function

LiverWeight

Converting CLuint to CLH


CLuint. MPPGL.Liver WeightCLuint =

MPPGL= 10 (1.407 +0.0158×age - 0.00038×age2 + 0.0000024×age3)

Liver Weight = Liver Volume × Liver Density

Liver Volume = 0.722.BSA 1.176 (L/m2)

0.00718×Ht 0.725×Wt 0.425

f (age)+x

(whole liver)

Converting CLuint to CLH


fuB =fu

CB/CpQH×fuB×CLuint

QH + fuB ×CLuintCLH =

CLuint×MPPGL×Liver WeightCLuint =

CB/Cp = (E:P)×HC + (1- HC)

HC=f (age)+f (sex)

QH = %CO

CO=f (age, BSA)

0.00718×Ht 0.725×Wt 0.425

f (age)+x


Revised in vivo ontogeny functions for CYP1A2 and 3A4(Leong et al., CPT 2012; 91: 926-931)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

0 5 10 15 20 25

Rela

tive

expr

essi

on

Age (y)

CYP1A2 ontogeny

in vitro

In vivo v14.1

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20 25

Rela

tive

expr

essi

on

Age (y)

CYP3A4 ontogeny

In vitro

In vivo v14.1


UGT Ontogeny

Strassburg et al 2002Burchell et al 1989Onishi et al 1997Leakey et al 1987Coughtrie et al 1988Miyagi and Collier 2007Zaya et al 2006Pacifici et al 1990Pacifici et al 1982Choonara et al 1989

Leiden Collaboration – Top down vs bottom up ontogeny for UGT2B7- Morphine - Zidovudine

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20

Frac

tion

of A

dult

Valu

e

Age (y)

UGT1A1

UGT1A4

UGT1A6

UGT1A9

UGT2B7


UGT2B7 ontogeny ‘Top down’ vs ‘Bottom up’

• Take home message is that pattern of ontogeny appears to be reasonable except for early neonates

• But under-prediction of CL across age band with morphine.

Bottom up

Top down

Bodyweight (kg) Bodyweight (kg)

Cle

aran

ce (L

/h)

Glu

curo

nida

tion

clea

ranc

e (L

/h)


Maturation of Renal Clearance

y = 87.674x - 14.497R 2 = 0.9988

0

50

100

150

0 0.5 1 1.5 2BSA (m2)

GFR

(ml/m

in)

921 subjects

Johnson et al 2006

0

20

40

60

80

100

120

140

160

0 5 10 15 20

GFR

(ml/

min

)

Age (yr)

Rhodin et al 2009Johnson et al 2006De Cock et al 2014Rubin et al 1949Hayton 2000

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140

Rhod

in, D

e Co

ck ,

Hayt

on (m

l/m

in)

Johnson (ml/min)

De CockRhodinHaytonLine of unity

923 subjects

63 subjects

1760 subjects


Maturation of Biliary Clearance Appears to be RapidAzithromycin Ceftriaxone

Digoxin Buprenorphine

Johnson et al Drug Metab Dispos. 2016


Variation in Protein Binding (fu)

Alb = 1.1287Ln(Age) + 33.746

0

10

20

30

40

50

60

0.1 1 10 100 1000 10000 100000

Age (days)

Albu

min

(g/L

)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0.1 1 10 100 1000 10000 100000Age (days)

AA

G (g

/L)

33.0D

38.0

38.0D

g/L Age89.8Age887.0AAG+

×=

DK]P[1

1fu+

= KD = Dissociation Constant[P] = Serum Protein Concentration 1

fu1

]P[KD

−=

( )

−×+

=

*

*

*

1][

][1

1

pop

pop

pop fufu

PP

fuIn absence of changes in dynamics of binding:

*pop is the population under investigation i.e paediatric


Developing and testing a Geriatric population

y = 0.0012x2 - 0.4357x + 196.38R² = 0.0475

0

20

40

60

80

100

120

140

160

180

200

50 75 100

Heig

ht (c

m)

Age (years)

Male - 66 to 96 y Height (males 66 to 96 y)

y = 478573x-1.346

0

500

1000

1500

2000

2500

60 70 80 90 100

Live

r wei

ght (

g)

Age (y)

Liver weight (Male)

Clinical

Simulated

Power (Simulated)

y = 6566.4x-0.792

R² = 1

0

100

200

300

400

60 70 80 90 100

Kidn

ey w

eigh

t (g)

Age (y)

Kidney weight (Female)

Clinical

Simulation

Power (Simulation)

0

10

20

30

40

50

60

70

60 70 80 90 100

Albu

min

(g/L

)

Age (y)

Albumin (males)

Simulated

Veering

Verbeeck

Campion

Parkinson et al 2004 CYPs

0

0.2

0.4

0.6

0.8

1

1.2

Elde

rly :

Youn

g C

L ra

tio

In house testing

Obs

Pred

http://www.google.co.uk/url?sa=i&rct=j&q=&esrc=s&frm=1&source=images&cd=&cad=rja&docid=fQ2w6O47GwQ52M&tbnid=KOoP8nkRh8vYwM:&ved=0CAUQjRw&url=http://pharmacologycorner.com/pharmacology-elderly-pharmacokinetics-polypharmacy-pp/&ei=ZN1CUu_kHqKV0AWfnYH4Bg&bvm=bv.53077864,d.d2k&psig=AFQjCNGFNTs620RkkQifIRl6BWV5RV0lIQ&ust=1380200120166102

http://www.google.co.uk/url?sa=i&rct=j&q=&esrc=s&frm=1&source=images&cd=&cad=rja&docid=fQ2w6O47GwQ52M&tbnid=KOoP8nkRh8vYwM:&ved=0CAUQjRw&url=http://pharmacologycorner.com/pharmacology-elderly-pharmacokinetics-polypharmacy-pp/&ei=ZN1CUu_kHqKV0AWfnYH4Bg&bvm=bv.53077864,d.d2k&psig=AFQjCNGFNTs620RkkQifIRl6BWV5RV0lIQ&ust=1380200120166102

Scaling from in vitro: drug data vs systems data

© Copyright 2013 Certara, L.P. All rights re

Jmax/Km orCLuint T

SF 1:

H-BMvIn vitroCLuint, T

CLuint, Tper

g Brain

SF 2: SF 3:

CLuint, Tper Brain @ BBB

REF/RAFH-BMv H-BMvPGB Brain Weight

BrainIn vitro data

Jmax/Km orCLuint T

SF 1:

HHEPIn vitroCLuint, T

CLuint, Tper

g Liver

SF 2: SF 3:

CLuint, Tper Liver

REF/RAFHHEP HPGL Liver Weight

LiverIn vitro data

served.

In vitro dataJmax/Km or

CLuint T

SF 1:

PTCIn vitroCLuint, T

CLuint, Tper

g Kidney

SF 2: SF 3:

CLuint, Tper Kidney

REF/RAFPTC PTCPGK Kidney Weight

Kidney

Caco-2, MDCK- II, LLC-PK1 etc.Jmax/Km or

CLuint T

SF 1:

CLuint, T

In Jejunum I

Intestine

REF/RAFJejunum I

Scaling via the Permeability and

Surface area product

Replacement / Additional Organ

Jmax/Km orCLuint T

CLu, Tper whole

organ

User needs to scale to whole organ!

SF: Scaling Factor


Translating in vitro effective concentrations to concentrations at the site of action

• Mechanistic, multi-compartmental tissue models (brain, kidney, liver, lung

and intestine) are available

• Enable more reliable estimates of intracellular tissue concentrations

22


Modelling in vitro assays – a must to do!

2 Compartment Model

Intracellular Volume

Free [S]EC

Incubation Medium Volume

Active UptakeCLint · Km · [S]EC(t)

Km + [S]EC(t)

CLPD· [S]EC(t)

CLPD · [S]IC(t)

[S]IC

fucell

Baker et al., Xenobiotica, 2007; Soars et al., Mol Phar, 2009; Poirier et al., Mol Pharm , 2009; Menochet et al., J Pharm Exp Ther, 2012

5 Compartment Model - Transwell

Heikkinen et al., 2010 Mol Pharmaceutics Korzekwa et al., 2012 DMD


PBPK Impact on 19 US Drug Labels in Last 2 Years

Olysio (Simerprevir) Xarelto (Rivaroxaban) Edurant (Rilpivirine) Imbruvia (Ibrutinib) Opsumit (Macitentan) Hepatitis C Thrombosis & Embolism HIV infection Lymphoma and Leukemia Pulmonary Hypertension

Zykadia (Ceritinbi) Odozmzo (Sonidegib) Farydak (Panobinostat) Revatio (Sildenafil) Bosulif (Bosutinib)Lung Cancer Basal Cell Carcinoma Multiple myeloma Pulmonary Hypertension Myelogenous Leukemia

Lynparza (Olaparib) Movantik (Naloxegol) Tagrisso (Osimertinib) Iclusig (Ponatinib) Cerdelga (Eliglustat)Advanced Ovarian Cancer Opioid Induced Constipation Metastatic NSCLC Chronic Myeloid Leukemia Gaucher Disease

Jevtana (Cabazitaxel) Cotellic (Cobimetinib) Lenvima (Lenvatinib) Aristada (Aripiprazolel) Prostate Cancer Metastatic Melanoma Thyroid cancer Schizophrenia


Quantitative IVIVE of Tissue Toxicity Supported by European Commission 7th FP Predict-IV Grant

25Hamon et al., Toxicology in Vitro, 2015


Summary

• In a systems pharmacology paradigm, the bottom-up approach tomodeling and simulation of the ADME processes of a chemical, is avaluable tool in integrating available prior information and improvingdecision making.

• Improvement in the in vitro systems which can act as surrogates for invivo reactions relevant to ADME

• Advances in the understanding of the extrapolation factors

• Advances in the development of mechanistic models of the humanbody

• Facilitate predicting PK characteristics in a wide range of healthy ordisease populations accounting for age, sex, ethnicity, genetic, etcvariability

• Moving towards PBPK coupling with systems biology models to predicttoxicity endpoints/biomarkers and their associated variability from invitro data

26

in vitro in vivo extrapolation and its applications in predicting pk … · 2020-06-29 · james r....

Documents