Mechanism-based PK-PD model of remoxipride with rat-to-human extrapolation: Mechanism-based PK-PD model of remoxipride with rat-to-human extrapolation: Mechanism-based PK-PD model of remoxipride with rat-to-human extrapolation:

characterizing remoxipride CNS target site PK and positive systems homeostatic feedback characterizing remoxipride CNS target site PK and positive systems homeostatic feedback characterizing remoxipride CNS target site PK and positive systems homeostatic feedback characterizing remoxipride CNS target site PK and positive systems homeostatic feedback

Jasper Stevens1, Bart Ploeger2, M. Hammarlund-Udenaes3, Piet Hein van der Graaf4, Meindert Danhof1, Elizabeth C.M. de Lange1Jasper Stevens1, Bart Ploeger2, M. Hammarlund-Udenaes3, Piet Hein van der Graaf4, Meindert Danhof1, Elizabeth C.M. de Lange11 Division of Pharmacology, LACDR, Leiden, NL, 2 LAP&P Consultants, Leiden, NL, Uppsala University, SE 4Pfizer Global R&D, Sandwich, Kent, UK1 Division of Pharmacology, LACDR, Leiden, NL, 2 LAP&P Consultants, Leiden, NL, Uppsala University, SE 4Pfizer Global R&D, Sandwich, Kent, UKDivision of Pharmacology, LACDR, Leiden, NL, LAP&P Consultants, Leiden, NL, Uppsala University, SE Pfizer Global R&D, Sandwich, Kent, UK

Results Pharmacokinetics cont.)Introduction Results Pharmacokinetics cont.)Introduction Results Pharmacokinetics cont.)Introduction

Our ultimate aim is to develop a translational MB PKPD model for dopaminergic

Results Pharmacokinetics cont.)

Our ultimate aim is to develop a translational MB PKPD model for dopaminergic plasma Brain-ECF

Our ultimate aim is to develop a translational MB PKPD model for dopaminergic

compounds, for the prediction of PKPD relationships following different routes of plasma Brain-ECFcompounds, for the prediction of PKPD relationships following different routes of plasma Brain-ECFcompounds, for the prediction of PKPD relationships following different routes of

administration and animal-human extrapolation. To that end, information on drug administration and animal-human extrapolation. To that end, information on drug

specific- and biological system specific parameters is needed. Many dopaminergic IN data

specific- and biological system specific parameters is needed. Many dopaminergic IN data

specific- and biological system specific parameters is needed. Many dopaminergic

compounds suffer from poor oral bioavailability and/or poor blood-brain barrier Fig 1. VPC for 1000 IN data

compounds suffer from poor oral bioavailability and/or poor blood-brain barrier Fig 1. VPC for 1000

simulations, depicted per compounds suffer from poor oral bioavailability and/or poor blood-brain barrier

transport. Intranasal administration could offer an attractive alternative route of simulations, depicted per

data set and per transport. Intranasal administration could offer an attractive alternative route of

administration by enhancing the distribution to the brain. Here we present a

data set and per

compartment. Badministration by enhancing the distribution to the brain. Here we present a

compartment. B

lue dots; observed REM administration by enhancing the distribution to the brain. Here we present a

translational PKPD model of prolactin (PRL) release following intravenous (IV) and lue dots; observed REM

concentrations over time,

IV datatranslational PKPD model of prolactin (PRL) release following intravenous (IV) and concentrations over time,

black line; median of IV datatranslational PKPD model of prolactin (PRL) release following intravenous (IV) and

intranasal (IN) administration of the dopamine D2-receptor antagonist remoxipride black line; median of IV data

intranasal (IN) administration of the dopamine D2-receptor antagonist remoxipride

(REM) as model compound. The model allows translation between i) different

black line; median of

simulations, grey area’s;

(REM) as model compound. The model allows translation between i) different simulations, grey area’s;

95% confidence intervals.(REM) as model compound. The model allows translation between i) different

routes of administration in rats, and ii) rat and human.

95% confidence intervals.

routes of administration in rats, and ii) rat and human.routes of administration in rats, and ii) rat and human.

MethodsMethodsResults Pharmacodynamics

MethodsResults Pharmacodynamics

Animal experiments (as previously publishedA). Dataset 1; 0, 4, 8, or 16 mg/kg REM Results Pharmacodynamics

Animal experiments (as previously publishedA). Dataset 1; 0, 4, 8, or 16 mg/kg REM Results Pharmacodynamics

Animal experiments (as previously published ). Dataset 1; 0, 4, 8, or 16 mg/kg REM

administration (30-min IV / 1-min IN infusion). Dataset 2; consecutive dosing of PRL profiles were accurately described following IV single as well as administration (30-min IV / 1-min IN infusion). Dataset 2; consecutive dosing of PRL profiles were accurately described following IV single as well as

consecutive dosing (fig 2 & 3, respectively). administration (30-min IV / 1-min IN infusion). Dataset 2; consecutive dosing of

3.8 mg/kg REM ( 30 min IV infusions, different time intervals). consecutive dosing (fig 2 & 3, respectively). 3.8 mg/kg REM ( 30 min IV infusions, different time intervals).

Analytical procedures. REM concentrations were analyzed by online solid phase

consecutive dosing (fig 2 & 3, respectively).

Analytical procedures. REM concentrations were analyzed by online solid phase Analytical procedures. REM concentrations were analyzed by online solid phase

extraction coupled to LC-MSB. PRL was measured in plasma (ELISA).extraction coupled to LC-MSB. PRL was measured in plasma (ELISA).extraction coupled to LC-MS . PRL was measured in plasma (ELISA).

Modeling. Modeling (NONMEM VI.2) was based on plasma- and brainECF REM Modeling. Modeling (NONMEM VI.2) was based on plasma- and brainECF REM

concentrations, and plasma PRL concentrations. PK was fixed during PD modeling.ECF

concentrations, and plasma PRL concentrations. PK was fixed during PD modeling.concentrations, and plasma PRL concentrations. PK was fixed during PD modeling.

Validation. Models were tested and evaluated (likelihood ratio test (p<0.05), Validation. Models were tested and evaluated (likelihood ratio test (p<0.05), Validation. Models were tested and evaluated (likelihood ratio test (p<0.05),

goodness-of-fit, parameter estimate endpoints, and individual plots).goodness-of-fit, parameter estimate endpoints, and individual plots).

Rat-to-human translation. Based on the animal PKPD model, human PKPD was Rat-to-human translation. Based on the animal PKPD model, human PKPD was Rat-to-human translation. Based on the animal PKPD model, human PKPD was

simulated in Berkeley-Madonna and compared to a clinical dataset, kindly provided simulated in Berkeley-Madonna and compared to a clinical dataset, kindly provided simulated in Berkeley-Madonna and compared to a clinical dataset, kindly provided

by M. Hammarlund-UdenaesC. Simulated human brainECF concentration effect-by M. Hammarlund-UdenaesC. Simulated human brainECF concentration effect-

relations were applied. Clinical PD parameters were obtained from literature when ECF

relations were applied. Clinical PD parameters were obtained from literature when relations were applied. Clinical PD parameters were obtained from literature when

possible, otherwise allometric scaling was applied. possible, otherwise allometric scaling was applied. possible, otherwise allometric scaling was applied.

ModelModelFig 2; VPC of PRL plasma profiles after single IV REM, Fig 3; typical individual predictions (black line) and ModelFig 2; VPC of PRL plasma profiles after single IV REM,

per dose (DO, in mg/kg). (O) observed PK (black line);

Fig 3; typical individual predictions (black line) and

observed PRL plasma PK O) after consecutive

REM PK-model. Following IN administration, total absorption of REM is defined by per dose (DO, in mg/kg). (O) observed PK (black line);

median of simulations, (grey area) and 95% confidence

observed PRL plasma PK O) after consecutive

dosing of REM, per dosing regiment (DOSE, hours)REM PK-model. Following IN administration, total absorption of REM is defined by median of simulations, (grey area) and 95% confidence

interv.als)

dosing of REM, per dosing regiment (DOSE, hours)REM PK-model. Following IN administration, total absorption of REM is defined by

systemic uptake (1.ABS) and direct nose-to-brain transport (2.ABS). REM is interv.als)

systemic uptake (1.ABS) and direct nose-to-brain transport (2.ABS). REM is

distributed (blue arrows) in a central-, peripheral- and brain compartment, and Results Translationdistributed (blue arrows) in a central-, peripheral- and brain compartment, and Results Translationdistributed (blue arrows) in a central-, peripheral- and brain compartment, and

eliminated from plasma (k ) and brain ECF (k ).

Results Translationeliminated from plasma (ke,REM,pl) and brain ECF (ke,REM,brainECF). The translational properties for IN REM indicate slight under-prediction eliminated from plasma (ke,REM,pl) and brain ECF (ke,REM,brainECF).

PK-PD model. An Emax model for relation between brainECF REM concentrations The translational properties for IN REM indicate slight under-prediction

of measurements (fig 4), attributed to larger variability in the IN dataset PK-PD model. An Emax model for relation between brainECF REM concentrations

and PRL release. A turnover model described PRL synthesis rate (k ), release of of measurements (fig 4), attributed to larger variability in the IN dataset

and PRL release. A turnover model described PRL synthesis rate (ks,PRL,rat), release of of measurements (fig 4), attributed to larger variability in the IN dataset

and more complex brain distribution. The model showed accurate and PRL release. A turnover model described PRL synthesis rate (ks,PRL,rat), release of

PRL from lactotrophs into plasma (k ), and PRL plasma elimination (k )and more complex brain distribution. The model showed accurate

PRL from lactotrophs into plasma (kr,PRL,rat), and PRL plasma elimination (ke,PRL,rat)and more complex brain distribution. The model showed accurate

translational properties in reduction of human PKPD (Fig 5, PK not PRL from lactotrophs into plasma (kr,PRL,rat), and PRL plasma elimination (ke,PRL,rat)

System-feedback model. An Emax model for the relation between PRL plasma translational properties in reduction of human PKPD (Fig 5, PK not

shown)System-feedback model. An Emax model for the relation between PRL plasma

concentration and rate of synthesis. shown)

concentration and rate of synthesis. k

shown)concentration and rate of synthesis.

ke, REM,plke, REM,pl

1.ABS 3.central 5.Peripheral1.ABS 3.central 5.Peripheral1.ABS 3.central 5.Peripheral

ph

arm

aco

kin

eti

csp

ha

rma

cok

ine

tics

ph

arm

aco

kin

eti

csp

ha

rma

cok

ine

tics

ph

arm

aco

kin

eti

cs

ke,REM, brainECF4.brain2.ABS

ph

arm

aco

kin

eti

cs

ke,REM, brainECF4.brain2.ABS

ph

arm

aco

kin

eti

csp

ha

rma

cok

ine

tics

ph

arm

aco

kin

eti

cs

Emax,REM,brainECF * CREM,brainECF(t) / (EC50,REM,brainECF + CREM,brainECF(t))Drug-effect Emax,REM,brainECF * CREM,brainECF(t) / (EC50,REM,brainECF + CREM,brainECF(t))Drug-effect

modelmodel

ph

arm

aco

dy

na

mic

s

kKK

ph

arm

aco

dy

na

mic

s

Fig 4 (up); translation to intranasal administration. VPC ke,PRL, ratKr, PRL, ratKs, PRL, rat

ph

arm

aco

dy

na

mic

s

TurnoverFig 4 (up); translation to intranasal administration. VPC

showing the median simulated PRL concentrations

lactotroph PRL, plasma

ke,PRL, ratKr, PRL, ratKs, PRL, rat

ph

arm

aco

dy

na

mic

s

Turnover

modelshowing the median simulated PRL concentrations

(black line) with the 95% confidence interval, and the lactotroph PRL, plasma

ph

arm

aco

dy

na

mic

s

model (black line) with the 95% confidence interval, and the

experimental data (open symbols)

ph

arm

aco

dy

na

mic

s

experimental data (open symbols)

ph

arm

aco

dy

na

mic

s

System-effect Fig 5 (right); translation to human PD. Based on

ph

arm

aco

dy

na

mic

s

System-effect

model

Fig 5 (right); translation to human PD. Based on

estimated human brainECF PK the predicted PRL PK (grey

1+ (SE * C (t) / (SEC + C (t))

ph

arm

aco

dy

na

mic

s

model estimated human brainECF PK the predicted PRL PK (grey

line) are in accordance with the clinical data (O)1+ (SEmax,PRL,rat * CPRL,pl, rat(t) / (SEC50,PRL,rat + CPRL,pl, rat(t))

ph

arm

aco

dy

na

mic

s

line) are in accordance with the clinical data (O)1+ (SEmax,PRL,rat * CPRL,pl, rat(t) / (SEC50,PRL,rat + CPRL,pl, rat(t))

Conclusions and perspectivesConclusions and perspectivesConclusions and perspectivesResults Pharmacokinetics

The preclinical PKPD model provides, for the1st time, quantification of Results Pharmacokinetics

The preclinical PKPD model provides, for the1st time, quantification of Results Pharmacokinetics

The preclinical PKPD model provides, for the1 time, quantification of

direct nose-to-brain transport in terms of rate and extent following IN Following IN administration total bioavailability was 89%. Brain distribution differs direct nose-to-brain transport in terms of rate and extent following IN

administration. It also indicates a positive feedback of PRL plasma

Following IN administration total bioavailability was 89%. Brain distribution differs

between administration routes. The brain ECF/plasma ratio improved for IN versus administration. It also indicates a positive feedback of PRL plasma between administration routes. The brain ECF/plasma ratio improved for IN versus administration. It also indicates a positive feedback of PRL plasma

concentrations on PRL synthesis. Finally it provides a useful approach for

between administration routes. The brain ECF/plasma ratio improved for IN versus

IV administration (28% vs. 19%). Brain concentrations were slightly under- concentrations on PRL synthesis. Finally it provides a useful approach for IV administration (28% vs. 19%). BrainECF concentrations were slightly under- concentrations on PRL synthesis. Finally it provides a useful approach for

rat-to-human translation of PKPD relationships of dopaminergic drugs.

IV administration (28% vs. 19%). Brain concentrations were slightly under-

predicted, indicating more complex brain distribution (fig 1) rat-to-human translation of PKPD relationships of dopaminergic drugs.predicted, indicating more complex brain distribution (fig 1)

A. Stevens et al., Pharm.Res,2009,26(8);1911-1917A. Stevens et al., Pharm.Res,2009,26(8);1911-1917

B. Stevens et al., J. Chromatogr. B Analyt Technol. Biomed. Life Sci., 2010, 878; 969-975 B. Stevens et al., J. Chromatogr. B Analyt Technol. Biomed. Life Sci., 2010, 878; 969-975 B. Stevens et al., J. Chromatogr. B Analyt Technol. Biomed. Life Sci., 2010, 878; 969-975

C. Movin-Osswald et al., J.Pharmacol.Exp.Ther.,1995, 274:921-927C. Movin-Osswald et al., J.Pharmacol.Exp.Ther.,1995, 274:921-927