The 2012 FDA Draft Guidance on Drug-drug Interactions:Drug drug Interactions:

Enzyme Induction and Beyond

Michael Sinz, Ph.D.

Research FellowResearch and DevelopmentResearch and DevelopmentBristol-Myers Squibb Wallingford, CT USA

NJDMDG Symposium April 26, 2012

The “Guidance for Industry”

Reaction phenotypingDDI (inhibition & induction)

In vivo clinical studieso c ca stud esLabeling

CYP3A4 mRNATranscription

Promoter (CYP3A4) Gene (CYP3A4)

rug CYP3A4

( ) ( )

Con

c. o

f Dr

loss ofefficacy

Drug Drug-OH

Slide 2

Time (hr)

C efficacy

General Scheme of Model-based Prediction: Investigational Drug Interacting with CYP Enzymes

CYP inhibition(reversible and time-dependent)

CYP induction

• Measure mRNA change by investigational • Measure enzyme activity in g y gdrug in cultured human hepatocytes from 3 or more donors• Estimate DDI parameters

I i i RNA d fi d th h ld?

Measure enzyme activity in human liver microsomes• Estimate DDI parameters

Basic Models

Is increase in mRNA > a predefined threshold?Or

Is the calculated R value < 0.9?R3 = 1/(1 + d x Emax x [I] / (EC50 + [I]))

Is the calculated R value >1.1Reversible inhibition, R1=1+[I]/Ki

TDE, R2=(Kobs+Kdeg) & Kobs=kinactx[I]/(KI+[I])

Is AUCR>1.25 (inhibition) or <0.8 (induction)Mechanistic Static Model

1 1Mechanistic

OR, Dynamic Model (PBPK)

AUCR= 1

(AgxBgxCg)x(1-Fg)+Fg

1

(AhxBhxCh)x Fm+(1-Fm)xModels

Slide 3

Conduct a clinical study using an appropriate probe substrate

Yes

Basic Induction Model

1 + d x Emax x [I] 1

R3 = Is increase in mRNA > a predefined threshold?

or

d = 1

EC50 + [I]or

Is the calculated R<0.9

[I] = maximum total (free+unbound) systemic concentration

E i h i i d iEmax is the maximum induction responseEC50 is the concentration causing half maximal effect

What if due to solubility or cytotoxicity issues th E d EC t b d t i d?

AUC/F2 Method ?

Options ?

Basic equation is designed to eliminate false negatives,

the Emax and EC50 can not be determined? Method ?

Slide 4

q g g ,but unfortunately leads to increased false positives,hence the need to move to the mechanistic models

Mechanistic Static Induction Model

AUCR = 1

(AgxBgxCg)x(1-Fg)+Fg

1(AhxBhxCh)xFm+(1-Fm)x( g g g) ( g) g ( h h h) m ( m)

Gut Liver

C = 1 + d x Emax x [I]

[I] + EC50

[I]g = Fa x Ka x Dose/Qen

[I]h = fu x (Cmax + (Fa x Ka x Dose/Qh))

A – reversible inhibitionB – irreversible inhibitiond = 1 (?), likely to be < 1 in the mechanistic model (0.3-0.8; ~0.5)d 1 (?), likely to be 1 in the mechanistic model (0.3 0.8; 0.5)Fa = 1 (if unknown)Ka = 0.1/min (if unknown), a value of 0.03/min is probably more reasonableFg – fraction available after intestinal metabolismF f ti f t i l f b t t

Slide 5

Fh – fraction of systemic clearance of substrate

Interpretation: Is AUCR>1.25 (inhibition) or <0.8 (induction)

Basic to Mechanistic Model

d = 1[I] = 2.3 uM

Basic Model

R 0 12 InductionEmax = 25EC50 = 5.3 uM

R3 = 0.12(R3< 0.9 implies induction)

InductionExpected

d = 0.5Fa = 1Fa 1Ka = 0.03/minDose = 81 uM[I] = 2.3 uMF 0 012

Mechanistic Static Model

AUCR = 0 21InductionE t dFu = 0.012

Emax = 25EC50 = 5.3 uMKdeg,g = 0.00032/min

(AUCR<0.8 implies induction)AUCR = 0.21 Expected

‘Moderate Inducer’

Slide 6

deg,gKdeg,h = 0.00048/minFg,h (midazolam)

Model System for Evaluating Induction

• Primary cultures of human hepatocytes (fresh or cryopreserved)– What about immortalized hepatocytes or NHR assays?

“At present, data generated from other in vitro systems are considered complementary and may be reviewed along with data

d i h l d h ”generated with cultured hepatocyte systems.”– EMA, “very well justified” immortalized cell lines acceptable

• Evaluate 3+ donors to account for interindividual• Evaluate 3+ donors to account for interindividual variability– If one donor is positive in Basic model, the drug is considered an

inducer and “follow-up evaluation is needed” (ie. Mechanistic)

• Need to determine performance of hepatocytes in identifying enzyme induction potential with ‘sufficient’identifying enzyme induction potential with sufficient number of clinical inducers (how many?)

Slide 7

Model System for Evaluating Induction

• “The changes in the mRNA level of the target gene should be used as an endpoint”– Not consistent with EMA guidance (activity & RNA)– Activity should be an option in cases when the test compound is

not an inhibitornot an inhibitor

• Inclusion of negative controls– Could provide value in distinguishing real induction from the ‘noise’ p g g

of the assay, but no guidance given on what to use or at what concentration.

• Evaluate CYP1A2 2B6 and 3A• Evaluate CYP1A2, 2B6, and 3A– If CYP3A is positive, then evaluate CYP2Cs (2C8, 2C9, and 2C19)

– Generally, CYP2C induction is less than CYP3A and mediated by factors be ond PXR and CAR (Fahmi et al DMD 2010 & Chen/Goldstein CDM 2009)beyond PXR and CAR (Fahmi et al, DMD 2010 & Chen/Goldstein, CDM 2009)

Slide 8

In Vitro CYP Inducers (Table 2)

CYP In vitro inducer as positive controls

Recommended concentration (uM) of

positive controls

Reported fold induction in

enzyme activityp y y1A2 Omeprazole

Lansoprazole25-100

1014-24

102B6 Phenobarbital 500-1000 5-10

(CITCO-EMA) (<100 nM)2C8 Rifampin 10 2-42C9 Rifampin 10 42C9 Rifampin 10 42C19 Rifampin 10 202D6 None identified3A4 Rif i 10 50 4 313A4 Rifampin 10-50 4-31

No information on ‘reported fold induction in mRNA response’

Slide 9

Classification of In Vivo Inducers (Table 4)CYP Strong Inducers

>80% decrease in AUC

Moderate Inducers50-80% decrease

in AUC

Weak Inducers20-50% decrease

in AUC1A2 Montelukast, phenytoin,

smokersMoricizine, omeprazole,

phenobarbital2B6 Efavirenz, rifampin Nevirapine

2C8 Rif i2C8 Rifampin

2C9 Carbamazepine, rifampin Aprepitant, bosentan, phenobarbital, St. John’s

wortwort2C19 Rifampin Artemisinin

3A4 Avasimibe, carbamazepine,

Bosentan, efavirenz,etravirine, modafinil,

Amprenavir, aprepitant, armodafinil, p ,

phenytoin,rifampin, St. John’s wort

, ,nafcillin

,clobazamechinacea,

pioglitazone, prednisone, rufinamide, vemurafenib

2D6 None known None known None known

Slide 10

2D6 None known None known None known

Change in Clinical Section

“… it is critical to evaluate the time it takes for the enzyme activities to return to normal when induction or TDI is involved so that

a third crossover period in which the interacting drug is removed”a third crossover period in which the interacting drug is removed”

osur

e

baseline

recovery period

MD

Z ex

po

induced

recovery period

Time (days)

Slide 11

……AND BEYOND

A Novel Mechanism of Enzyme Induction:Non Classical PXR or CAR-Mediated Induction

CYP3A4 Gene Activation by PXR and CAR

CYP3A4 mRNATranscription

D

PXR DPXR

RIFHPF

RXR

TranslationBinding

Promoter (CYP3A4) Gene (CYP3A4)PXR

CARD

RXR

RXR

D D OH

CYP3A4

CAR

DPHTNVP

ug

Drug Drug-OHCAR

Both PXR and CAR can cause

onc.

of D

ru

loss of

Both PXR and CAR can cause CYP3A4 gene activation andenzyme induction leading

to significant drug interactions

Slide 13Time (hr)

Co efficacy

to significant drug interactions

Comparative PXR Binding and Transactivation Results

PXR Binding Assay

BMS-536924 binds to PXR: BMS-665351 does not bind to PXR:BMS-536924 binds to PXR: IC50 = 1.0 uM

BMS-665351 does not bind to PXR: IC50 = 30 uM

SAR

se e

PXR Transactivation Assay

nt R

espo

ns

t Res

pons

EC50=5 uM

Perc

en

Perc

en

Slide 14

Conc. uM Conc. uM

PXR and Hepatocyte Data on Both Sets of Compounds

Binds to PXRActivates PXR

BMS-536924Do Not Bind to PXRDo Not Activate PXR

BMS-665351

Activates PXR

Induces CYP3A4 in PrimaryHuman Hepatocytes

Do Not Activate PXR

Induces CYP3A4 in PrimaryHuman HepatocytesHuman Hepatocytes Human Hepatocytes

Slide 15

Rifampicin-positive control for PXR

Two Mechanisms of CAR-Mediated Enzyme Induction

CA

R

D

ConstitutiveAndrostane

Receptor Mechanism 2

CAR

R

D

PB

CARD

CAR CYP3A4

Mechanism 1

CAR

CARD

CYP3A4

CYP3A4

CYP3A4CYP3A4mRNA and

proteinnucleus

Transit of CAR from the

Slide 16

PB-phenobarbital

Transit of CAR from the cytoplasm to the nucleus by any mechanism is termed Translocation

Activation of CAR and PXR in HepG2 Cells

HepG24hCAR3/CYP3A4 (PXRE/XREM)

HepG26

CAR3 Expression Assay PXR Expression Assay

uc A

ctiv

ity

2

3

hCAR3/CYP3A4 (PXRE/XREM)

uc A

ctiv

ity

3

4

5hPXR/CYP3A4 (PXRE/XREM)

** **

Rel

ativ

e Lu

1

2

Rel

ativ

e Lu

1

2

3

0.1%DMSOCITCO 1uM 1uM 5uM

0

0.1% DMSORIF 10uM 1uM 5uM

0

BMS-665351BMS-665351

BMS-665351 does not activate human CAR or PXR

Slide 17

CITCO-positive control for CAR

CAR Translocation Assay: BMS-665351

Human hepatocytes infected with adenovirusfluorescently tagged with CAR (AD/EYFP-hCAR)

CTL (0.1%DMSO) PB 1 mM hCAR1 Localization in HL-#37100 C

ount

(%)

60

80

00 CC=NN

BMS-665351 (1 µM) BMS-665351 (5 µM) Rel

ativ

e C

20

40

CTL PB 1mM 1uM 5uM0

BMS-665351

C-cytoplasmN-nucleus

BMS-665351 does not translocate CAR

Does the Induction Have Anything to do with CAR?

CAR3 Over Expression System (HepG2)on

NA

expr

essi

oYP

3A4

mR

NC

CAR is involved in the CYP3A4i d ti BMS 665351 10 M

Slide 19Li, Sinz, et al. JPET Dec. 2011

PK11195 – a selective CAR deactivator

induction response BMS-665351: 10 uMCITCO: 1 uMPK11195: 10 uM

Does BMS-665351 Induce the Expression of CAR?tio

n 6

7HepG2Huh7

tion

4

5HL-#16HL-#17

****

**

Human Hepatocytes

NA

Fold

Indu

ct

3

4

5

NA

Fold

Indu

ct

2

3

**

***

**

hCAR

mR

N

1

2

3

hCAR

mR

N

1

2

*

0.1% DMSORIF 10uM

CITCO 1uM 1uM 5uM0

0.1% DMSO RIF 10uM

PB 1mM 1uM 5uM0

BMS-665351 BMS-665351

BMS-665351 induces the expression of CARin cell lines and human hepatocytes

D t i d th i f PXRDoes not induce the expression of PXR

Summary

CYP3A4 mRNATranscription

PXR

GC

VDR

TranslationPromoter (CYP3A4) Gene (CYP3A4)D

CARRXR

D D OH

CYP3A4

CARCAR ?

BMS-665351IGF1R Inhibitor Drug Drug-OH

CAR?

IGF1R Inhibitor

• Further studies necessary to link the increase in CAR expression to the increased expression of CYP3A4– CAR promoter-reporter assay and siRNA CAR knock-down– Does this mechanism of CYP3A4 induction translate to an in vivo DDI?

Slide 21

Acknowledgements

Bristol Myers Squibb University of BaltimoreBristol Myers SquibbSean KimMark Wittman

University of BaltimoreLinhao LiHongbing Wang

Kurt ZimmermannJoan CarboniBMS IGF1R TeamBMS IGF1R Team

Slide 22Slide 22