Transcript The Role of Drug Metabolism Studies in Optimizing Drug Candidates

The Role of Drug Metabolism Studies in
Optimizing Drug Candidates
Kenneth Santone, PhD
Bristol-Myers Squibb
Metabolism and Pharmacokinetics /
Pharmaceutical Candidate Optimization
ALTERNATE TITLE:
Why All the Chemist's Wonderful
Compounds Don't Become Drugs!
Our Focus
 Unmet medical need
 First in class
 Best in class
 Need for efficiency and
productivity enhancement
What are we faced with?
 Industrialization of pharmaceutical research
– Unprecedented increase in identification of targets
– Corresponding increase in throughput of chemistry
– Blurring of traditional discovery-development interface
 Focus and emphasis on “developability”
(early go/no go decisions)
 Improve success rate
 Reduce development timeline
– Necessity for increasing efficiency and productivity
Drug Discovery Paradigm Shift
‘Old’ Model
of Drug Discovery
Hits
Design
& Synthesis
Efficacy &
Selectivity Testing
‘New’ Model
of Drug Discovery
Validated Hits
Design
Efficacy &
& Synthesis
Selectivity
Testing
PAT
Screening &
Predictions
Lead Candidates
Physicochemical, ADME
& Tox Workup
Development Compound
Detailed Physicochemical,
ADME &Tox Workup
Development Compound
More informed
decision making
during Lead
Optimization,
through quicker and
earlier evaluation of
PAT attributes
The Hand-off from Drug Discovery to Development:
The Top Ten Quotations We All Know and Love*
10. “The molecular weight? 850. Why? Is that a problem?”
9. “We’ll need eight different capsule strengths for Phase I.”
8. “The compound is very potent in the in vitro screen but does not work well in the
animal efficacy model.”
7. “Now that you mention it, our solutions were a little cloudy.”
6. “The compound is highly insoluble but Pharmaceutical Development will fix the
problem.”
5. “BMS-XXXXXX is a highly potent and selective inhibitor of (the target).
In mouse models, the optimal dose was 200 mg/kg.”
4. “Toxicity?! It’s not the drug; must be a metabolite unique to that animal species.”
3. “Animal bioavailability ranged from 65% to <1%, depending on species.”
2. “Gee, we didn’t have any problems when we gave it in DMSO.”
1. “It’s a great compound, but it has formulation problems.”
Partially adapted from R.A. Lipper
*why great compounds don’t always become drugs
Critical Interfaces in Drug Discovery*
Chemistry
Biology
Activity
Safety
Metabolism & Pharmacokinetics
Pharmaceutics
Optimized Compound
*Analytical Chemistry (Bioanalysis) involved
in every one of these disciplines
Role of ADME* Studies
 Selection of quality drug candidate for development
– Developability
– First-in-class vs. best-in-class
– Crisp go/no go decisions
 Optimization of drug discovery and early development processes
– Multi-tiered approach for ADME studies
– Equal partnership with all functional areas
Lead Discovery
Biology
Chemistry
Pharmaceutics
Drug Safety
Analytical R&D
Clinical Pharmacology Process Chemistry
 Blurring of traditional discovery-development interface
* Absorption, Distribution, Metabolism, Excretion
Selection of Drug Candidates:
Focus on Developability
 Permeability
 Protein binding
 Transport
 Biopharmaceutics
 Metabolic stability
 Active/reactive/
 P-450 mediated drug
interactions
 PK/PD assessment
 Distribution
toxic metabolites
 In vivo
PK/bioavailability
in animals
 Prediction of PK
and efficacious
doses in humans
Tiered-Approach for ADME Studies
Hits to Lead
• In vitro Studies
•Permeability
•P450 inhibition
•Metabolic Stability
•In silico predictions
•Objective
•Develop SAR
•Chemotype selection
Tiered-Approach for ADME Studies
Lead Optimization
In vitro Studies
•Permeability/transport
•P450 inhibition
•Metabolic Stability
•Reaction phenotyping
•Protein binding
•In vivo PK
•Cassette dosing
•Individual PK
•Tissue penetration
•Early biotransformation
•Objective
•Identify a lead compound
•Feedback to chemistry/biology
Tiered-Approach for ADME Studies
Lead Selection
• Absolute bioavailability in pharmacology/toxicology models
• Dose dependency in PK
• Mechanism of absorption
• Assess potential for DDI
• Characterization of metabolites, routes of elimination
• Assess formation of active metabolites
• Interspecies differences in metabolism and in vitro-in vivo correlation
• Extrapolation of ADME properties to man from in vitro and in vivo
data
• Determination of PK/PD relationships; help selection of doses for First
in Human studies
•Objective
•Characterize the lead compound
•Identify risks/opportunities
How In Vitro Metabolic Stability Relates to Clearance?
TBC = CLhepatic + CLrenal + CLother
CLhepatic = CLmetabolism + CLbiliary
CLmetabolic = fB * CLintrinsic * Qh / fB * CLintrinsic + Qh
well stirred model of organ extraction
Intrinsic Clearance (CLi) = Vmax / Km = vo / Cu
through rearrangement of the Michaelis-Menton eqn, assuming drug conc is < Km
Depletion or Half-Life Method:
CLi = (0.693 * liver wt) / (in vitro t1/2 * amount of liver)
Tools to Predict Metabolic Clearance
In Vitro Systems
 Liver microsomes
– high throughput and most common
– mostly oxidative (CYP & FMO)
 S9 fraction
– high throughput
– Phase I & Phase II metabolism
In Vivo Animal Clearance
In Silico
 Hepatocytes
–
–
–
–
low throughput
cell membrane/transporters
intracellular concentration
Phase I & Phase II metabolism
In Vitro - In Vivo
Correlation
Metabolic Stability to Select Compounds with
Potentially Longer Half-Life
Human Metabolic Stability: Microsome vs Hepatocyte
Rate of Human Microsomal Metabolism
Microsome Total Metabolic Rate
0.4
0.40
0.30
0.20
0.2
BMS:Y
0.1
0.0
-1
0.10
1
2
3
4
BMS-231975
BMS-451491
BMS-233406
BMS-250598
BMS-436039
BMS-428028
BMS-221970
BMS-437220
BMS-437917
BMS-229983
BMS-275816
BMS-212435
BMS-437562
BMS-440883
BMS-451503
BMS-214662
BMS-225975
BMS-214662
BMS-437221
BMS-201282
BMS-212347
BMS-271494
BMS-225263
BMS-227178
BMS-338387
BMS-437134
BMS-434841
BMS-435689
BMS-227178
0.00
0
Hepatocyte Metabolic Rate
BMS-350869
Rate of metabolism
Oxidation
Glucuronidation
2
R = 0.8
0.3
• Lead compound is primarily glucuronidated in humans
• Human in vitro systems with combination of oxidation and
glucuronidation employed for selection of back up
5
Major Reactions Involved in Drug Metabolism
OXIDATIVE REACTIONS (CYP, LM+NADPH)
•N-Dealkylation: erythromycin, morphine, caffeine
•O-Dealkylation: codeine, dextromethorphan
•Aliphatic Hydroxylation: tolbutamide, midazolam
•Aromatic Hydroxylation: phenytoin, amphetamine, warfarin
•N-Oxidation: chlorpheniramine, dapsone
•S-Oxidation: cimetidine, omeprazole
•Deamination: amphetamine
Major Reactions Involved in Drug Metabolism
HYDROLYSIS REACTIONS (Esterase, ?LM+NADPH)
•Ester Hydrolysis: aspirin, cocaine
•Amide Hydrolysis: lidocaine, procainamide
CONJUGATION REACTIONS (Phase II, hepatocytes)
•Glucuronidation: morphine, ibuprofen
•Sulfation: acetaminophen
•Acetylation: sulfonamides, isoniazid
Metabolic Stability Summary
•
Not all metabolism is hepatic.
•
Incubation concentration < Km balanced with assay sensitivity.
•
Need to correlate with in vivo model.
•
Fast in vitro clearance generally implies fast in vivo clearance, the reverse need not be
true.
•
Confounding physical-chemical properties.
 solubility, stability, purity, non-specific binding
•
Real concentration at enzyme active site?
 protein binding, cell penetration, non-specific binding
•
In vitro systems generally underestimate CLi due to non-specific binding.
•
Can the stability be too good? Yes, in certain situations.
•
Many unknown factors to can contribute to a poor in vitro - in vivo correlation or poor
estimation of human metabolic stability.
 Nonetheless, in vitro methods are still the best method for predictions
Drug-Drug Interaction Summary
• Major drug interactions are caused by either inhibition or induction of
drug metabolizing enzymes.
• Semi-quantitative predictions of drug interactions
 many unknown factors
 human ADME properties in vivo
• Models provide numbers that must be placed in context with multiple
factors:
 therapeutic area
 therapeutic index, route of administration
 market competition
• Animal models are not predictive of human interaction potential ???
• Static nature of in vitro systems compared to the dynamic in vivo system
• Mixtures of interaction mechanisms from the same compound are
extremely difficult to predict:
 reversible + irreversible inhibition
 inhibition + induction
Assessment of Active Metabolites
Compound
Met Ratea
Cmaxb
(M)
AUC(0-8)b
(M.h)
IC50
(nM)
BMS-X
BMS-Y
0.64
0.58
8.5
13.2
35
64.4
19
19
Efficaceous
Dose
(mole/kg)
1.4
>60
a: metabolism rate in nmol/min.mg protein in rat liver microsomes
b: rat oral exposure studies at 0.1 mmol/kg
Issue
• Similar metabolism and in vitro activity profile but different in vivo
activity profile
• Apparent PK/PD disconnect
Solution
• Rapid in vitro metabolism and biological activity assays
Assessment of Active Metabolites
Compound
BMS-X
BMS-Y
In vitro Activity of Liver Microsomal
Product in Cell Based Assay (IC50 (nM))
Parent
0 min
30 min
incubation
incubation
19
12
19
19
60
490
% parent
remaining
<1
20
Structural identification of active metabolites
• MS/MS indicated presence of monohydroxylation
• NMR showed site of hydroxylation
Subsequent steps
• Monohydroxylated metabolite synthesized
• Activity and PK properties confirmed
Assessment of Reactive Metabolites
•A number of functional (chemical) elements have been
associated with problems in drug discovery leading to toxicity
 Metabolic activation to reactive intermediates
 Interference with metabolic processes
•Clinical manifestations include (preclinical measure)
 Cellular (hepatic) necrosis (animal toxicity)
 Idiosyncratic toxicity (glutathione adducts, protein
covalent binding, immunogenic response)
 Drug-drug interactions (mechanism-dependent CYP
inhibition)
Examples of Reactive Metabolites
Furans
O
O
O
CYP3A4
(epoxidation)
O
O
O
O
O
CYP3A4
(epoxidation)
O
OH
O
O
O
OH
OCH3
Aflatoxin B
6',7'-dihydroxybergamottin
Furan substructure is associated with toxicity (eg. aflatoxin) and
with CYP inhibition (eg. bergamottin)
Examples of Reactive Metabolites
Thiophenes
S
O
O
S
S
Nu
O
CYP2C9
O
S
HO
H2N
O
O
S
N
O
O
Cl
Cl
Tienilic acid
Cl
Tenidap
Thiophene substructure has been associated with several types of
toxicity (predominately hepatotoxicity). Other thiophene
containing drugs: ticlopidine, clopidigrel, raloxifene.
Examples of Reactive Metabolites
Anilines, Nitroaromatics
NH2
NO2
HN
OH
O
O
H2N
N
H
O
O
S
O
S
N
N
Sulfamethoxazole
O
H2N
NH2
Dapsone
Anilines are associated with a number of types of toxicity (eg. methemoglobinemia, skin
rashes, etc.). Nitroaromatics are primarily activated by initial reduction, often in the
gut, followed by N-oxidation.
Anilines of polycyclic aromatic systems are often potent mutagens and carcinogens (eg.,
naphthylamine, aminofluorene) through conjugation of the hydroxylamine and
subsequent loss of the conjugate to leave a nitrenium ion.
Examples of Reactive Metabolites
Amines, alkylamines
O
N
N
O
S
O
N
O
O
N
Diltiazem
The metabolism of amines or alkylamines is generally related to timedependent inhibition of CYP enzymes, with the nitroso species forming a tight
complex with the heme iron, known as a MI complex. Other compounds that
undergo this type of transformation and inhibit CYPs are TAO, erythromycin
and verapamil
Examples of Reactive Metabolites
Quinone, Quinoid
O
O
X = O, N, C
X
X
O
O
HN
O
O
S
NH
HO
O
OH
Acetaminophen
Troglitazone
Quinone-like compounds can exert their effects through direct
alkylation of nucleophiles or through redox cycling between their
oxidized and reduced forms
Examples of Reactive Metabolites
Acetylenes
O
N
OH
O
OH
O
Gestodene
Mifepristone (RU 486)
Acetylenes have been found to be time-dependent inhibitors of
CYP enzymes.
Examples of Reactive Metabolites
Acyl glucuronidation formation
Direct reaction with nucleophiles
O
O
OH
OGluc
Amidori rearrangement, then reaction with nucleophiles
O
O
N
N
O
O
Cl
Zomipirac
OH
OH
Tolmentin
Acyl glucuronides have been implicated in both direct hepatic
damage and idiosyncratic toxicities
Challenges and Opportunities
 HTS screens for prediction of permeability, metabolic stability, metabolic
reactivity and DDI
– How are we using these data?
– Retrospective analysis on return of investment
– The numbers in gray zone!
– Secondary assays for better predictability
 Application of animal PK/bioavailability data for lead optimization
– Adequacy of permeability and metabolic stability data
– Animals vs. humans: quantitative and qualitative differences in ADME
properties
 Informed decision based on drug metabolism and pharmacokinetic data
– Low bioavailability vs. oral efficacy
– Role of metabolite(s), reactivity of metabolite(s)
– Protein binding
– In vitro- in vivo correlation in animals and extrapolation to humans
 Issue of enzyme induction in humans
– In-vitro models and predictability
– False and real alarm from in-vivo animal data
Challenges and Opportunities
 Use of biomarkers
– In-vivo biology, animals vs. humans
– Development and validation of assays
– Transfer from preclinical to clinical laboratories
– Biomarkers = Surrogate marker = Efficacy/Toxicity
– A balancing act of emerging science
 The feedback loops
– To and from chemistry
– To and from biology
– To and from drug safety
– To and from pharmaceutics
– To and from clinical pharmacology
 Volume of data
– Conversion of information into knowledge
– Timing and availability
A Focused Application of ADME Studies
• Active involvement earlier in the Discovery Process
• Timely guidance to Chemistry to select chemotypes with
desirable ADME properties
• Maximize informed decision making during Lead
Optimization
• Improved ability to predict human metabolism and
pharmacokinetics
• Stronger partnerships with Drug Discovery and all areas
of Pharmaceutical Development
Our Mission
To ensure that no development candidate
fails in the clinic due to an
unforeseen metabolic or
pharmacokinetic property
Acknowledgements
David Rodrigues and Griff Humphreys
Saeho Chong, Punit Marathe, Wen Chyi Shyu
and Mike Sinz
And finally ….