PBPK modeling to characterize the interplay between metabolism and

transport in the disposition of simeprevir in healthy volunteers and

HCV infected patients

Sivi Ouwerkerk-Mahadevan1 and Jan Snoeys1

1Janssen Research & Development, Beerse, Belgium

Simeprevir (SMV/TMC435)

Phase III trials of SMV+PR in HCV GT 1- and GT 4-infected treatment-naïves and relapsers showed SVR12 rates ~80%1–4

In a Phase II trial (COSMOS) of SMV + sofosbuvir ± ribavirin in HCV GT 1-infected treatment-naïves and prior null responders, the overall SVR12 rate was 92.2%5,6

Safe and well tolerated (~3800 patients treated in clinical trials to date) In development as part of IFN-free combinations

GT, genotype; IFN, interferon; PR, peginterferon-α-2a/-2b + ribavirin; SVR12, sustained virologic response 12 weeks after end of treatment

1. Jacobson I et al. Lancet In press; 2. Manns M et al. Lancet In press; 3. Forns X et al. Gastroenterology Mar 3 [Epub ahead of print];

4. Moreno C et al. HepDART 2013; 5. Lawitz E et al. EASL 2014; 6. Sukowski MS et al. EASL 2014.

O O O O N H

N H

O O

N N

N

S

MeO

(S)

(R) (R) (R) (R)

S

Once-daily capsule, HCV NS3/4A protease inhibitor

Approved in Japan, Canada, the USA, Russia and Europe

Antiviral activity in patients infected with HCV GT 1, 2, 4, 5, and 6

Substantial interpatient variability in plasma concentrations – No correlation with sustained virologic response (SVR) at therapeutic dose

More than dose-proportional increase in exposure

Increase in exposure on repeated dosing – Steady-state achieved in approximately 7 days

Difference in pharmacokinetics between healthy volunteers and patients: 2–3-fold higher exposure in patients

Ethnic differences: higher exposure in Japanese patients

Substrate of several uptake and efflux transporters

Pharmacokinetic challenges of simeprevir

Dose-normalized AUC vs dose in HCV patients

CV, coefficient of variation

Phase 2/3 studies C201, C205, C206, C208, C216, HPC3007

Dose (mg)

% CV N

25 52 18

75 79 177

100 87 196

150 111 1133

200 75 30

2000

1500

1000

500

0 0 50 100 150 200

SMV dose (mg)

SMV

dose

-nor

mal

ized

AU

C

(ng.

h/m

L/m

g)

AUC in non-Asian global phase 3 studies and in a Japanese phase 2 study

500 400

C215 Asian n=38

AUC

(µg.

h/m

L)

300 200

100

150 mg 100 mg

Non-Asian Asian

Phase 3 non-Asian

n=757

Phase 3 Asian n=14

Reason for non-linear PK

Reason for large interpatient variability

Mechanism for difference between healthy volunteers and HCV-infected patients

Ethnic differences

Drug–drug interactions in HCV-infected patients

Use of PBPK modeling to understand:

PBPK, physiologically based pharmacokinetic

PBPK model components

Drug-dependent component

System component (drug-independent)

Intestines

Lung

Oral dose

Rapidly perfused organs

Slowly perfused organs

Kidney

Liver

ADME, PK, PD, and MOA Metabolism Active transport Passive diffusion Protein binding Drug–drug interactions Receptor binding

PBPK model Dosing

Elimination

MOA, mode of action

Simeprevir parameters used for simulations

Physicochemical properties

Low solubilities at all pH

Saturation of hepatic uptake (non-linear distribution)

Saturation of liver and gut CYP3A4 (non-linear clearance)

Clinical pharmacokinetics in healthy volunteers

Not included: – Active efflux in gastro-intestinal tract via, for example, P-gp; not

considered to play a major role, confirmed by data from drug-drug interaction with cyclosporine and data from single dose ritonavir

Model verification: Simulating the known – single dose studies

Simulation of healthy Caucasians (N=50) Observed data

Minimal change in half-life despite pronounced non-linear increase in exposure with dose – Saturable gut and liver CYP3A4 metabolism, and saturation of active uptake

Simultaneous decrease in both Vd and Cl, therefore t1/2 remains constant

100000

10000

1000

100

10

1.0

0.1 8 0 16 24 32 40 48 56 64 72

Time (h)

600 mg 450 mg 300 mg 200 mg

150 mg 100 mg 50 mg

Plas

ma

conc

entra

tion

(mg/

mL)

100000

10000

1000

100

10

1 0 12 24 36 48 60 72

Time (h)

600 mg 450 mg 300 mg 200 mg

100 mg 50 mg

Plas

ma

conc

entra

tion

(ng/

mL)

Model verification: Simulating the known – single dose studies

AUC dose normalized versus the 50 mg SMV dose. Simulations performed in 54 healthy Caucasian male volunteers. Observed data in 6 patients except for 150 mg SMV dose in 24 patients

Simulated average (±SD) and observed average SMV plasma AUC0–72h after single dose administration D

ose

norm

aliz

ed A

UC

(ng/

mL)

SMV dose (mg)

35000

0 0 50 100 150 200 250 300 350 400 450 500 550 600

5000

10000

15000

20000

25000

30000

Simulated Observed

Model verification: Simulating the known – multiple dose studies

Dose and time-dependant PK is explained by saturable clearance and saturable distribution – Lack of time-dependent inhibition was also confirmed from midazolam DDI studies

100000

10000

1000

100

10

1.0

0 24 0 48 72 96 120 144 168

Time (h)

350 mg QD 200 mg QD 150 mg QD 100 mg QD

Plas

ma

conc

entra

tion

(ng/

mL)

140000

120000

80000

60000

20000

0 50 100 150 200 250 300 350

SMV dose

Dos

e no

rmal

ized

mea

n SM

V AU

C

(ng/

ml.h

)

40000

100000

Simulated Observed

Simulation of the following interactions, and comparison with clinical DDI studies in healthy volunteers: – CYP3A4 interaction without impact on hepatic uptake

● Ritonavir (100 mg QD/BID) and erythromycin (500 mg TID) inhibit intestinal and hepatic CYP3A4 and intestinal P-gp but not hepatic OATPs

– OATP interaction without impact on metabolic clearance ● Cyclosporine (100 mg) inhibits intestinal P-gp and hepatic OATPs

– Combined OATP and CYP3A4 interactions ● Rifampicin (600 mg QD) inhibits OATP and induces liver and intestinal

CYP3A. Also inhibits hepatic MRP2 and induces intestinal P-gp

Model verification: DDI studies

MRP, multidrug resistance-associated protein; P-gp, P-glycoprotein

Efavirenz 600 mg QD – modification of published model by Rekic et al. fugut was reduced to remove

intestinal CYP3A4 induction

Ritonavir 100 mg QD and BID – modified to include CYP3A4 inactivation kinetic parameters so that complete

inhibition of gut and liver CYP3A4 is achieved (Ernest et al., Mathias et al.)

Rifampicin 600 mg QD – Model by Xu et al. was used. Induction of CYP3A4 on Day 1 was removed.

OATP1B1/3 Ki modified based on published statin interactions (Maeda et al.)

Erythromycin 500 mg TID – Simcyp Sim-file was used unmodified

Cyclosporine 100 mg – Model by Varma et al. was modified. CYP3A4 Clint added. Lag time, Fa, Ka included

in first-order absorption model, muscle Kp reduced to match observed Vd,ss. OATP inhibition, Fu, inc and CYP3A4 Ki were modified to match published DDI data

Model verification: Compound file modification to match clinical data

Model verification: Ritonavir DDI

Plasma concentration time profile for 200 mg QD. SMV administration for 7 days with and without 100 mg BID ritonavir in healthy Caucasian volunteers (study C104), and simulated data in 36 patients

Simulated GMR Observed GMR

Cmax AUC Cmax AUC

Ritonavir 5.8 10 4.7 (3.8–5.8) 7.2 (5.6–9.2)

28000

0 24

Time-substrate (h)

Plas

ma

conc

entra

tion

(ng/

mL)

24000

20000

16000

12000

8000

4000

20 16 12 8 4

CSys CSys with interaction 28000

0 0 24

Time-substrate (h)

Plas

ma

conc

entra

tion

(ng/

mL)

24000

20000

16000

12000

8000

4000

20 16 12 8 4

SMV 200 mg QD, day 7 (n=12) SMV 200 mg QD + Ritonavir 100 mg BID, Day 12 (n=12)

0

PBPK DDI summary Source: FDA clinical pharmacology review (modified)

PBPK model simulated and observed exposure changes of SMV by different enzyme and/or transport inhibitors and inducers

Inhibitor / inducer (mechanisms)

SMV dose

AUC (Cmax) ratio Explanation of observed DDI findings

Sim Obs

Ritonavir (Strong CYP3A inhibitor)

Single 2.1 (13) 1.8 (1.3) CYP3A inhibition augmented OATP saturation over time Multiple 10 (5.8) 7.2 (4.7)

Erythromycin (moderate CYP3A inhibitor)

Single 1.4 (1.1) - Erythromycin DDI study was done using lower SMV dose (150mg) than the ritonavir

study (200mg). At lower doses, stronger interaction is seen

Multiple 6.2 (3.7) 7.5 (4.5)

Cyclosporine (OATP inhibitor) Multiple 1.3

(Cmin ratio) 1.2

(Cmin ratio) OATP saturation after multiple dosing

diminished inhibitor effect

Rifampin (strong CYP3A inducer,

OATP inhibitor)

Single 2.1 (1.8) - OATP inhibition + CYP3A4 induction: Cmax and AUC of SMV Multiple 0.6 (1.1) 0.5 (1.3)

Efavirenz (moderate CYP3A inducer) Multiple 0.3 (0.6) 0.5 (0.3) CYP3A induction only, no effect on OATP

Cmax AUC

Summary

Figure reproduced with permission of John Keogh, JPK Consulting and modified

CYP3A4

Hepatocyte

Low doses

OATPs

CYP3A4

Hepatocyte

High and/or multiple doses

OATPs

Saturating

Non-linearity of SMV PK is described by Saturation of both OATP Uptake and CYP3A4 metabolism

CYP3A4

Hepatocyte

Ritonavir or Erythromycin Day 1

OATPs

CYP3A4

Hepatocyte

Ritonavir or Erythromycin Day 7

OATPs

Saturation

Saturation of OATP due to increased circulating Simprevir following inhibition of CYP3A4.

CYP3A4

Hepatocyte

Cyclosporine single dose at SS

OATPs

Near saturation of OATPs following repeat dose of victim accounts for relatively modest increase in exposure following single dose administration of a potent OATP inhibitor

CYP3A4

Hepatocyte

Rifampicin multiple dose

OATPs

CYP3A4

Hepatocyte

Efavirenz multiple dose

OATPs

Inhibition of OATPs by Rifampicin, but not by Efavirenz explains observation of increased Cmax by Rifampicin, despite decreased AUC observed for both CYP3A4 inducers

Already approaching Saturation

Cmax AUC

Changes were incorporated into the model for mild hepatic impairment Liver volume in HCV patients was 10% lower1,2 CYP3A4 abundance was decreased from 137 pmol/mg protein to

108.1 pmol/mg protein3

This resulted in 30% lower liver CYP3A4 Clint

A population of HCV patients was created

1. Lin et al Hepatogastroenterology 1998; 2. Johnson et al Clin Pharmacokinet 2010; 3. Nakai et al. DMD 2008

1 0

2 3 4 5 6 7 8

Non-HCV* F1 F2 F3

NS NS

p<0.05

CYP

3A4/

GAP

DH

1

0

2

3

4

5

6

Non-HCV* F1 F2 F3

NS p<0.05

p<0.01

OAT

P-C

/GAP

DH

Healthy volunteers

Simulating the known: Healthy volunteers and patients dosed SMV 150mg QD

0 31 62 93 124 155 186 217 248 279 310 0 31 62 93 124 155 186 217 248 279 310 0

1,000

2,000

3,000

4,000

5,000

Plas

ma

conc

entra

tion

(ng/

mL)

Time (h)

4,500

3,500

2,500

1,500

500

Patients

tmax

(h) Cmax

(ng/mL) AUC

(ng/mL.h)

3.90 2129 26951

Plas

ma

conc

entra

tion

(ng/

mL)

Time (h)

tmax

(h) Cmax

(ng/mL) AUC

(ng/mL.h)

4.52 4385 71316

0

1,000

2,000

3,000

4,000

5,000 4,500

3,500

2,500

1,500

500

Liver volume in Chinese patients is >30% smaller than in Caucasian patients Hepatic CYP3A4 abundance per mg protein is slightly lower in Chinese patients vs

Caucasian patients Lower OATP1B1 hepatic uptake transporter expression levels in Japanese patient

CYP3A abundance in Caucasian vs Chinese patients

Barter et al. Clin Pharmacokinet 2013; Tomita et al. Clin Pharmacol Ther 2013

Gut CYP3A Hepatic CYP3A

Caucasian Chinese

CYP

abu

ndan

ce (n

mol

)

0

500

1000

1500

2000

Live

r vol

ume

(mL)

Caucasian Chinese

Liver volume

137

116 120

99

020406080

100120140160

CPY3A4 CYP3A5

66

25

58

22

0

10

20

30

40

50

60

70

CPY3A4 CYP3A5

CYP

abu

ndan

ce (n

mol

)

Simulations: Chinese vs Caucasian healthy volunteers

Chinese patients (N=40)

Incorporating differences in liver volume and CYP3A4 expression into the model gives plasma exposures in Chinese patients 2-fold higher than in Caucasian patients at 100 mg

0 12 24 36 48 60 72 84 96 108 120 24 36 48 60 72 84 96 108 120 Time (h)

Upper CI

0 12

1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500

Plas

ma

conc

entra

tion

(ng/

mL)

Mean Lower CI

Cmax (ng/mL)

AUC (ng/mL.h)

2071 38292

Cmax (ng/mL)

AUC (ng/mL.h)

1098 17895

0 500

Time (h)

1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500

Plas

ma

conc

entra

tion

(ng/

mL)

0 500

Caucasian patients (N=40)

5,000 5,000

Simulating the unknown: Liver concentrations of SMV in Japanese and Caucasian patients

Physiological factors incorporated into the model: Japanese patients have 15% lower liver volume, lower CYP3A4 abundance, and intrinsically lower OATP1B1 baseline activity

Liver 150 mg Caucasian patients Liver 100 mg Japanese patients Plasma 150 mg Caucasian patients Plasma 100 mg Japanese patients

80000 SM

V co

ncen

tratio

n (n

g/m

l) 70000

60000

50000

40000

30000

20000

10000

0 460 456 464 468 472 476 480

Time (h)

Simulating the unknown: DDI in patients

AUC ratio healthy

volunteers

AUC ratio HCV patients

CYP 3A inhibitors

Ritonavir 100 mg BID 10 5.1

Erythromycin 6.2 4.9

CYP 3A inducers

Efavirenz 0.3 0.23

Rifampicin 0.54 0.32

OATP substrate

Rosuvastatin 2.9 2.8

Regulatory feedback FDA PBPK modeling review memo online

Questions addressed by the submitted PBPK modeling report and additional information requested by the Office of Clinical Pharmacology include: – What are the major mechanisms contributing to non-linear PK

of SMV? – Can DDI with SMV be predicted?

In addition, sponsor simulated PK of SMV in various specific populations and projected liver concentrations of SMV in Caucasian and Asian HCV patients

Regulatory feedback FDA PBPK modeling review memo online

Conclusion Sponsor’s PBPK modeling and simulation reasonably captured

non-linear PK of SMV Saturation of OATP transporter mediated drug distribution into

the liver and saturation of CYP3A4 metabolism together appear to be the plausible mechanisms contributing to the non-linear PK and differential effects of CYP3A4 and/or OATP modulators observed in the DDI studies

The model can be used to predict other untested DDI situations and to evaluate the effect of various intrinsic factors (eg. ethnicity, liver disease) on SMV exposure

Acknowledgements

The authors would like to thank the volunteers and:

– Jan Snoeys – Maarten Huisman – Anne Brochot – Maria Beumont-Mauviel – Alex Simion

Medical writing support was provided by Martin Goulding on behalf of Complete Medical Communications, funded by Janssen