simulating oral absorption with pk-sim methodology and ... · simulation of dose dependent exposure...

Simulating Oral Absorption with PK-Sim®

Methodology and Application Examples

Stefan Willmann, Andrea N. Edginton, Walter Schmitt, Marcus Kleine-Besten, and Jennifer B. Dressman

[email protected]

Presentation Overview

PK-Sim®’s Absorption Model

Application Examples

Example 1

Example 2

Example 3

Simulating Drug AbsorptionDr. Stefan Willmann, PAGE 2006

PK-Sim® Model-Structure

Real Integrated Whole Body Modelcomprising:

Fully integrated GI-tractBiliary tract, enables enterohepaticcyclingMost important organsFor each organ:

• metabolizing pathways• different active transporter types

(influx, efflux, Pgp-like)

Capability for treating even very sophisticated problems.

VEN

OU

S B

LOO

D P

OO

L

AR

TER

IAL

BLO

OD

PO

OL

PANCREAS

SPLEEN

PORTAL VEIN

LIVER

KIDNEY

LUNG

DOSEi.v.

p.o.

GALL BLADDER

STOMACH

WALL

SMALL INTESTINE

WALL

LARGE INTESTINE

WALL

TESTESHEART

BRAINFAT

BONESKIN

MUSCLE

VEN

OU

S B

LOO

D P

OO

L

AR

TER

IAL

BLO

OD

PO

OL

PANCREAS

SPLEEN

PORTAL VEIN

LIVER

KIDNEY

LUNG

DOSEi.v.

p.o.

GALL BLADDER

STOMACH

WALL

SMALL INTESTINE

WALL

LARGE INTESTINE

WALL

TESTESHEART

BRAINFAT

BONESKIN

MUSCLE

Willmann et al., Biosilico 1(4), 121-124 (2003)

PK-Sim®‘s Input Parameters

Lipophilicity (LogMA preferred)

Plasma Protein Binding

(alternatively unbound fraction)

(effective) Molweight

pKa values for acids/bases

Solubility vs. pH table

Plasma Clearance hepatic/renal

(alternatively in vitro metaboli-

sation rates, Km & Vmax, …)

Transit Function (for rat): Graphical Representation

Simulation of Intestinal Transit & Absorption

Continuous one-compartment model for small intestine:• Spatially varying properties (effective surface area, pH)• GI transit described as plug-flow with dispersion

Principle

Willmann et al., Pharm. Res. 20, 1766-1771 (2003)

Model for Passive Absorption

Model for the intestinalpermeability coefficient was build using a data set of 126marketed compounds with nosolubility limitation at thera-peutic doses.

An excellent fit was obtained.

All outliers are known to besubstrates active transporters.

Willmann et al., J. Med. Chem. 47, 4022-4031 (2004)

]s/cm[MWD

MWCMAMWBMW

MAMWA)MA,MW(P

arparacellullartranscellu

int

44 344 2144444 344444 21

γ−γ−

γ−

β−α−

β−α−

++

+= ]s/cm[

MWDMWC

MAMWBMWMAMWA)MA,MW(P

arparacellullartranscellu

int

44 344 2144444 344444 21

γ−γ−

γ−

β−α−

β−α−

++

+=

Intestinal permeability is based on an empirical equation*:

•D. Leahy et al. in Novel Drug Delivery and Its Therapeuthic Application (1989)




Example 1: Dissolution Limited Absorption

Example 2

Example 3


Compound X (from ongoing BHC development project)

• neutral• medium lipophilicity (LogMA = 2.2)• MW ~ 450 g/mol• high permeability (Caco2, animal models)• low aqueous solubility

• Human study data:

• single dose under fasted conditions • solution: 5 and 10 mg• IR tablet: 1.25, 5, 10, 15, 20, 30, 40, 60, and 80 mg

(mean particle diameter of IR tablet formulation: 3 µm)

Properties


Model for Dissolution-Limited Absorption

Add-on module to dynamically simulate dissolution of poly-disperse spherical particles


0 1 2 3 4 5 6 70 %

20 %

40 %

60 %

80 %

100 %

3 µm Particles

1.25mg 2.50mg 5.00mg 10.0mg 20.0mg 30.0mg 60.0mg 80.0mg

Frac

tion

Dis

solv

ed

Time [h]

Simulation of Dose Dependent Exposure

0 %

20 %

40 %

60 %

80 %

100 %

0 20 40 60 80 1000

2

4

6

8

10

Mea

sure

d AU

Cno

rm [k

g h

/ L]

Dose [mg]

Experiment Tablet Solution

Simulation Tablet Solution

Sim

ulat

ed F

abs

0 2 4 6 8 10 120

50

100

150

200

250

300

350

400

IR Tablet: 1.25mg 5.00mg 10.0mg 15.0mg 20.0mg 30.0mg 40.0mg 60.0mg 80.0mg

Pla

sma

Con

cent

ratio

n [µ

g/L]

Time [h]

0 2 4 6 8 10 120

100

200

300

400

Solution: 5.00mg 10.0mg

Pla

sma

Con

cent

ratio

n [µ

g/L]

Time [h]





Example 1

Example 2: Cimetidine PK Variability

Example 3


What About Inter-Individual Variability ?

?


Age Dependence and Variability of Physiologicaland Anatomical Parameters

Data for the age dependence of physiological parametersrelevant for PBPK modellingsuch as

- body weight, body height, body massindex,

- organ weights, blood flow rates,- tissue composition (water, lipid, and

protein content)- fractions vascular, interstitial and

intracellular

and their variability and cross-correlation were collected in a comprehensive literature search.

Modelling Inter-Individual Variability in PK-Sim®

with the add-on module “PK-Pop“


Available Information: - in vitro dissolution profiles of IRTagamet® tablets plus threeexperimental CR formulations(400 mg cimetidine)

- in vivo data from 12 malevolunteers

- PK profiles after iv admin. insame individuals (clearance distribution !)

Population Simulation:- 100 virtual individuals (age, weight and height matched) with varyinggastric emptying time (15–45 min.) and small intestinal transit time (2,5–

5,5 h)

Example: Cimetidine

0 1 2 3 4 5 6

0 %

20 %

40 %

60 %

80 %

100 %

TAGAMET EUDR75 EUDR15 EUDR26C

imet

idin

e R

elea

se

Time [h]

Exp. Data from: Jantratid et al., Clin. Pharmacokin. (2006;45(4):385-99)

FaSSIF @ pH 6.5


Example: Cimetidine

(unpublished data, Marcus Kleine-Besten, Univ. Frankfurt)

0 1 2 3 4 5 6

0 %

20 %

40 %

60 %

80 %

100 %


imet

idin

e R

elea

se

Time [h]

In vitro release profile:

Time [h]0 2 4 6 8 10 12

Pla

sma

Con

cent

ratio

n [m

g/L]

0

1

2

3

4

5

Mean5% and 95%Min/Max

Tagamet® tablets

Example: Cimetidine


0 1 2 3 4 5 6

0 %

20 %

40 %

60 %

80 %

100 %


imet

idin

e R

elea

se

Time [h]


Time [h]0 2 4 6 8 10 12

Pla

sma

Con

cent

ratio

n [m

g/L]

0

1

2

3

4

5

Mean

Min/Max5% and 95%

Eudragit 7,5 %

Example: Cimetidine


0 1 2 3 4 5 6

0 %

20 %

40 %

60 %

80 %

100 %


imet

idin

e R

elea

se

Time [h]


Time [h]0 2 4 6 8 10 12

Pla

sma

Con

cent

ratio

n [m

g/L]

0

1

2

3

4

5

Mean5% and 95%Min/Max

Eudragit 15 %

Example: Cimetidine


0 1 2 3 4 5 6

0 %

20 %

40 %

60 %

80 %

100 %


imet

idin

e R

elea

se

Time [h]


Time [h]0 2 4 6 8 10 12

Pla

sma

Con

cent

ratio

n [m

g/L]

0

1

2

3

4

5

Mean5% and 95% Min/Max

Eudragit 26 %




Example 1

Example 2

Example 3: Simulations in Children


Special Populations: Children

Food & Drug Administration (FDA) and European authorities pressuring industry to perform clinical trials in paediatric patients

Increase access to treatments for children and reduce off-label, unlicensed treatment

50% of all drugs administered in hospitals are not properly licensed for use in children*

Adverse drug reactions due to dosing errors and use in non-labeled ages - from Dec 2001-2004, 820 serious ADRs occurred due to off label use with 130 being fatal (reported)**

* Conroy et al. Survey of unlicensed and off label drug use in paediatric wards in European countries. British Medical Journal 320:79-82. (2000)** European Medicines Agency. Evidence of harm from off label or unlicensed medicines in children. EMEA/11207/04. (2004)

Scaling PBPK Models to Children

PBPK Modelling in children requires knowledge about the

physiological and clearance differences relative to adults.






Children are not just small adults!





-> mechanistic model for clearance scaling

(Edginton et al., Clin. Pharmacokin. (in press 2006))

Simulating Oral Absorption in Children

Example: Oral administration of 10 mg/kg Ciprofloxacin in

children


Simulating Drug Absorption: Summary

The rate and extent of oral drug absorption can be well

simulated based on simple physico-chemical input parameters

and a detailed description of the GI physiology

Physiology-based simulations of drug absorption can be used

to make predictions in the early development phase

for hypothesis testing in later development stages

to aid formulation development

to study sub-populations (e.g. children, elderly, diseases)


The PK-Sim®-TeamBTS-Systems Biology & Computational Solutions- Dr. Andrea Edginton- Karsten Höhn- Marcus Kleine-Besten- Dr. Jörg Lippert- Dr. Walter Schmitt- Michael Sevestre- Juri Solodenko- Wolfgang Weiss- Dr. Stefan Willmann

Bayer-internal Cooperations- PK groups of BHC-Pharma

(Dr. G. Ahr, Dr. W. Mück, Dr.H.Stass and colleagues)

External Cooperations- Prof. Jenny Dressman,

Uni Frankfurt- NIMBUS Biotechnology, Leipzig- Physiomics plc, UK


B A C K U P S L I D E S

Relevant Physiological Parameters: Summary

(Physiological data collected in collaboration with Prof. Dressman, Frankfurt, data for dogs and mice not shown)

human rat

Simulating Drug AbsorptionDr. Stefan Willmann, EDAN 2006

Transit Function (for rat): Graphical Representation

Simulation of Intestinal Transit & Absorption

Willmann et al., Pharm. Res. 20, 1766-1771 (2003)

substance-specific properties physiological properties

( ) )!S)t,z(C()z(A)]t(C)t,z(C[Pdzd

dtdz)t,z(Md

intlumeneffpvlumenintpv

2

≤−=

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20Age (y)

Rel

. Int

rinsi

c C

lear

ance

Sum

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20Age (y)

Rel

. Int

rinsi

c C

lear

ance

Scaling of Intrinsic Clearances

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 5 10 15 20

Age (y)

Rel

. Int

rinsi

c C

lear

ance

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

0 5 10 15 20

Age (y)

Rel

. Int

rinsi

c C

lear

ance

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 5 10 15 20

Age (y)

Rel

. Int

rinsi

c C

lear

ance

Renal

CYP 3A4

CYP 1A2

20 %

30 %

50 %

Hypothetical Compound

26 %

23 %

51 %

Validation of Clearance Scaling

R2 = 0.91R2 = 0.910.01

0.1

1

10

100

0.01 0.1 1 10 100

Training compounds Alfentanil Midazolam Gentamicin Isepamicin Caffeine Ropivacaine Morphine Lorazepam

Test compounds Buprenorphine Lidocaine Ciprofloxacin Paracetamol Theophylline Fentanyl

Observed Clearance [ml/min/kg]

Pre

dict

ed C

lear

ance

[ml/m

in/k

g]

One Pathway

Multiple Pathways0.01

0.1

1

10

100

0.01 0.1 1 10 100

R2 = 0.5246Q2 = 0.7977(all: 0.6137)

Clearance in Children [ml/min/kg]

Adu

lt C

lear

ance

[ml/m

in/k

g]

R2 = 0.61R2 = 0.61

Using adult clearances would lead to much worse prediction!

Calculation of Partition Coefficients

tissuewaterwaterprotein

tissueproteinwaterlipid

tissuelipidwatertissue fKfKfK +∗+∗= ///

Steady state organ/plasma partition coefficients

K = Partition Coefficient (Kprotein/water = HSA binding in case of plasma and calculated from Lipophilicity in all other cases)

f = Volume fraction

waterplasma

watertissue

plasma

tissueplasmatissue K

KCCK

/

// == watertissueK /= .fu

Calculation of Distribution Dynamic

Permeability x Surface-Area Products

Fick’s First Law: )( 21 CCAdDJ −⋅⋅= C2

C1

D

d

)( 21 CCAdDKJ −⋅⋅=

PAtissue ∼ Lipophilicity . MW-α . Atissue

Volume of Distribution

0 5 10 15 20 25 300

5

10

15

20

25

30

VD

ss m

easu

red

VDss calculated

Beta Blockers

Doxorubicin

Prediction of Partition Parameters

0.01

0.10

1

10

100

0.01 0.10 1.00 10.00 100.00

Measured

Cal

cula

ted

actively influxed

actively effluxed

Validation

Brain/Plasma Partition-CoefficientFraction Unbound in Plasma

0,0

0,2

0,4

0,6

0,8

1,0

0,0 0,2 0,4 0,6 0,8 1,0

Measured

Cal

cula

ted

0,0

0,2

0,4

0,6

0,8

1,0

0,0 0,2 0,4 0,6 0,8 1,0

Measured

Cal

cula

ted

RatHuman

Keldenich et al., oral presentation at LogP2004 Symposium, Zürich (2004)

Organ/Plasma Partition CoefficientsMore Examples

Willmann et al., Poster presentation at LogP2004 Symposium, Zürich (2004)

Membrane Affinity

Partitioning into Phospholipid-

Membrane

Chains

Headgroups

MA = Cmembrane / CwaterMA = Cmembrane / Cwater

Polar compound Lipophilic compound

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

+-

Cmembrane

Cwater

Membrane Affinity vs. logKow

Relationship logKow / logMA nonlinear

logMA (Mw)

pH-Dependence different from logKow

Example: Basic compound

Austin et al. , Fisons PharmaceuticalsI. logP Symposium, Lausanne 1995

Gobas et al. , J. Pharm. Sci 77, 265 (1988)

General Model Results

Assumption: Dissolution is not the rate limiting step for absorption


General Model Results

Region A: no solubility limitation

Region B: dose-dependentabsorption

Region C: Fabs < 1 %(solubility-limited)

Region D:Fabs < 1 %(permeability-limited)

FDA-Suggestion: (M/S) = 250 ml

“Minimum Acceptable Solubility“ *)

*) W. Curatolo (Pfizer) Pharm. Sci. Technol. Today 1(9), 387-393 (1998)


Example: Cimetidine

0

2

4

6

8

10

12

30 40 50 60 70 80 90 100 110 120 130 140 150

No.

Liver [ml/ min]

Plasma clearance

0

2

4

6

8

10

12

58 60 62 64 66 68 70 72 74 76 78 80

No.

kg

Weight

0

2

4

6

8

10

152 154 156 158 160 162 164 166 168 170 172 174 176

No.

cm

Height

0

2

4

6

8

180 200 220 240 260 280 300 320 340 360

No.

GI-Tract [min]

Intestina l T ransit T ime

0

2

4

6

8

10

0 10 20 30 40 50 60

No.

GI-Tract [min]

Gastric Emptying T ime


Model Application: Assessment of EHC

Example: Model compound:

weak base with pKa = 8.5,

LogMA = 4,2, MW = 520,

Solubility = 250 mg/L at pH 6.5,

dose = 25 mg, subject to EHC

0 2 4 6 8 10 121

10

100


Plas

ma

Con

cent

ratio

n [m

g/L]

Time [h]

GBE




LogMA = 4,2, MW = 520,



Population Simulation (N=25) 0 2 4 6 8 10 121

10

100

Pla

sma

Con

cent

ratio

n [m

g/L]

Time [h]

0

1

2

3

180 190 200 210 220 230 240 250 260 270 280 290 300

No.

min

EHC Lag T ime

0

1

2

3

4

500 600 700 800 900 1000 1100

No.

Liver [ml/ min]

Plasma clearance

0

1

2

3

4

500 600 700 800 900 1000 1100

No.

Gallbladder [ml/ min]

Plasma clearance




LogMA = 4,2, MW = 520,



Population Simulation (N=25)

0

1

2

3

180 190 200 210 220 230 240 250 260 270 280 290 300

No.

min

EHC Lag T ime

0

1

2

3

4

500 600 700 800 900 1000 1100

No.

Liver [ml/ min]

Plasma clearance

0

1

2

3

4

500 600 700 800 900 1000 1100

No.

Gallbladder [ml/ min]

Plasma clearance

0 1 2 3 4 5 6 7 8 9 10 11 121

10

100

Pla

sma

Con

cent

ratio

n [m

g/L]

Time [h]

simulating oral absorption with pk-sim methodology and ... · simulation of dose dependent exposure...

Documents