Metabolic Activation and Idiosycratic
Drug Toxicity: By Avoiding Structural
Alerts, Do We Mitigate Risks?
Amit S. Kalgutkar, Ph.D.
Pfizer Global Research and Development
Groton, CT 06340, USA
1
Cause(s) of Attrition in Drug Discovery
• In the early 90’s, major cause of attrition was
poor pharmacokinetics1,2
– Largely resolved via involvement of DM/PK groups at early stages of drug
discovery (Exploratory/Lead development/candidate-seeking)
• Of late: lack of efficacy (achieving POM for
novel targets) and drug safety are the leading
causes of candidate attrition
– Pharmacology tactics to counterbalance attrition
• Better understanding of pharmacological targets
• Incorporation of translational pharmacology (PK/PD, disease
biomarkers, etc)
• Probe concept (exploratory INDs, etc)
– Tactics to counterbalance safety-related attrition arising from IADRs
• ?
1Roberts SA
(2003) Drug metabolism and pharmacokinetics in drug discovery. Curr Opin Drug Discov Devel. 6(1):66-80.
2Kola I and Landis J (2004) Can the pharmaceutical industry reduce attrition rates? Nat Rev Drug Discov. 3(8):711-5
2
Safety-Related Attrition: Adverse Drug Reactions (ADRs)
• ADRs Contribute to patient morbidity and mortality
– One of the most common causes for drug recalls or black box warning labels
• Of a total of 548 drugs approved in the period from 1975-1999, 45 drugs (8.2%) acquired 1
or more black box warnings, 16 (2.9%) were withdrawn from the market
• ADR Classification
– Type A ADRs: ~ 80% of ADRs fall in this category
• Type A ADRs can be predicted from known drug pharmacology (e.g. Hemorrhage with
anticoagulants)
• Dose dependent – can be reversed with dose reduction
• Generally identified in preclinical species (animal models of pharmacology)
– Type B (Bizarre) or Idiosyncratic ADRs (e.g., hepatotoxicity, skin rashes, aplastic
anaemia, agranulocytosis) – e.g., black box warning for “sulfonamides”
• Unrelated to primary pharmacology
• Dose independent (with the exception of some drugs) – can occur at any dose within the
therapeutic range
• Temporal relationship - Symptoms subsides after cessation of treatment; rapid onset upon
re-challenge
• Can be severe - maybe fatal - most common cause for drug withdrawal
• Cannot be predicted from traditional toxicological studies in animals
• Rare - frequency of occurrence - 1 in 10,000 to 1 in 100,000
– Normally not observed until phase III or post launch
3
Associating a Functional Group with Adverse Drug Reactions
– Sulfonamides have achieved notoriety with respect to
hypersensitivity (e.g., skin rashes)
See - Kalgutkar, A. S., Jones, R. and Sawant, A. ”Sulfonamide as an Essential Functional Group in Drug Design” In Metabolism, Pharmacokinetics and
Toxicity of Functional Groups: Impact of the Building Blocks of Medicinal Chemistry on ADMET (Royal Society of Chemistry),
[Dennis A. Smith, Editor], 2010, Chapter 5, pp. 210 – 273.
4
BLACK BOX WARNING
FATALITIES ASSOCIATED WITH THE ADMINISTRATION OF SULFONAMIDES, ALTHOUGH RARE,
HAVE OCCURRED DUE TO SEVERE REACTIONS, INCLUDING STEVENS-JOHNSON SYNDROME,
TOXIC EPIDERMAL NECROLYSIS, FULMINANT HEPATIC NECROSIS, AGRANULOCYTOSIS,
APLASTIC ANEMIA, AND OTHER BLOOD DYSCRASIAS. SULFONAMIDES, INCLUDING SULFONAMIDE
CONTAINING PRODUCTS SUCH AS TRIMETHOPRIM/SULFAMETHOXAZOLE, SHOULD BE
DISCONTINUED AT THE FIRSTAPPEARANCE OF SKIN RASH OR ANY SIGN OF ADVERSE REACTION. In
rare instances, a skin rash may be followed by a more severe reaction, such as Stevens-Johnson
syndrome, toxic epidermal necrolysis, hepatic necrosis, and serious blood disorder (see PRECAUTIONS
). Clinical signs, such as rash, sore throat, fever, arthralgia, pallor, purpura, or jaundice may be early
indications of serious reactions.
N
CH3
O O
S
N
H
NH2
N
NH2
O
N
H3CO
H2N
OCH3
OCH3
Sulfamethoxazole
(Bactrim®)
Trimethoprim
5
6
The Concept of Xenobiotic Bioactivation to
Reactive Metabolites (RMs)
Origins in the field of chemical carcinogenicity
Ames Test for genotoxicity has S-9/NADPH-dependent bioactivation arm; required for FDA
submissions
• RM “covalently adducts” to DNA resulting in genotoxic response
•
– Fungal mycotoxin aflatoxin B1 (AFB1) – established hepatocarcinogen†
• Exposure occurs primarily through ingestion of mold-contaminated foods (e.g.,
corn and peanuts)
Rate-limiting step is P450-catalyzed RM formation
O
O
O
OCH3
O
OCH3
P450
H
Reaction with DNA-base(s) Guanine - (DNA adducts)
Reaction with glutathione
Reacttion with water
H
O
O
H
AFB1
O
O
O
H
Furan epoxide (A reactive metabolite)
†Guengerich FP, Johnson WW, Shimada T, Ueng YF, Yamazaki H and Langouet S (1998) Mutat. Res. 402:121-128.
§Miller EC and Miller JA (1947) Cancer Res. 7:468-480.
7
RMs and CYP Isozyme Inactivation
• RM “covalently adducts” to metabolizing enzymes (e.g.,
cytochrome P450) responsible for its formation
– Leads to enzyme inactivation and:
• Non-linear PK if P450 enzyme is also primarily responsible for clearance
• Drug-drug interactions (DDIs) (Atorvastatin/Grapefruit juice)
– Furanocoumarins, bergamottin and 6’,7’-dihydroxybergamottin, the abundant
constituents of GFJ, are mechanism-based inactivators of P4503A4
R
R
O
O
O
O
P4503A4
P4503A4 Inactivation
O
O
O
OH
R = H2C
R = H2C
OH
He K, Iyer KR, Hayes RN, Sinz MW, Woolf TF and Hollenberg PF (1998) Chem. Res. Toxicol. 11:252-259.
Kent UM, Lin HL, Noon KR, Harris DL and Hollenberg PF (2006) J. Pharmacol. Exp. Ther. 318:992-1005.
8
Toxic Drug Metabolites – Acetaminophen as an
Example
• Brodie et al. (National Institutes of Health) first to demonstrate:
– Bioactivation of acetaminophen and covalent binding to liver tissue
• Nelson et al. elucidated the mechanism of acetaminophen bioactivation
(involving a 2 electron oxidation to a reactive quinone-imine
intermediate)
– “Gold standard” of human and animal hepatotoxicity assessments
O
O
HN
N
CH3
 Dose-dependent (> 1 gm/day) hepatotoxin
O
CH3
HN
GSH
P450
O
S
O
OH
Acetaminophen
 Depletes GSH upon toxic overdose
 Covalent binding to > 30 hepatic proteins
 liver toxicity can be observed in animals
 N-Acetylcysteine as antidote
CH3
OH
O
NH
N
H
COOH
COOH
NAPQI
NH2
Reactive quinone-imine
GSH = glutathione – endogenous antioxidant (~ 10 mM concn. in mammals)
14C-APAP
covalent binding to microsomes prevented by GSH; confirms the protective role of the thiol
9
Drugs Associated With IADRs
Drugs Withdrawn
Aclcofenac (antiinflammatory)
Hepatitis, rash
Alpidem (anxiolytic)
Hepatitis (fatal)
Amodiaquine (antimalarial)
Hepatitis, agranulocytosis
Amineptine (antidepressant)
Hepatitis, cutaneous ADRs
Benoxaprofen (antiinflammatory)
Hepatitis, cutaneous ADRs
Bromfenac (antiinflammatory)
Hepatitis (fatal)
Carbutamide (antidiabetic)
Bone marrow toxicity
Ibufenac (antiinflammatory)
Hepatitis (fatal)
Iproniazid (antidepressant)
Hepatitis (fatal)
Metiamide (antiulcer)
Bone marrow toxicity
Nomifensine (antidepressant)
Hepatitis (fatal), anaemia
Practolol (antiarrhythmic)
Severe cutaneous ADRs
Remoxipride (antipsychotic)
Aplastic anaemia
Sudoxicam (antiinflammatory)
Hepatitis (fatal)
Tienilic Acid (diuretic)
Hepatitis (fatal)
Tolrestat (antidiabetic)
Hepatitis (fatal)
Troglitazone (antidiabetic)
Hepatitis (fatal)
Zomepirac (antiinflammatory)
Hepatitis, cutaneous ADRs
Temp. Withdrawn
or Withdrawn in
other Countries
Aminopyrine (analgesic)
Agranulocytosis
Nefazodone (antidepressant)
Hepatitis (> 200 deaths)
Trovan (antibacterial)
Hepatitis
Zileuton (antiasthma)
Hepatitis
Marketed Drugs
Abacavir (antiretroviral)
Cutaneous ADRs
Acetaminophen (analgesic)
Hepatitis (fatal)
Captopril (antihypertensive)
Cutaneous ADRs, agranulocytosis
Carbamazepine (anticonvulsant)
Hepatitis, agranulocytosis
Clozapine (antipsychotic)
Agranulocytosis
Cyclophosphamide (anticancer)
Agranulocytosis, cutaneous ADRs
Dapsone (antibacterial)
Agranulocytosis, cutaneous ADRs,
aplastic anaemia
Diclofenac (antiinflammatory)
Hepatitis
Felbamate (anticonvulsant)
Hepatitis (fatal), aplastic anaemia
(fatal), severe restriction in use
Furosemide (diurectic)
Agranulocytosis, cutaneous ADRs,
aplastic anaemia
Halothane (anesthetic)
Hepatitis
Imipramine (antidepressant)
Hepatitis
Indomethacin (antiinflammatory)
Hepatitis
Isoniazid (antibacterial)
Hepatitis (can be fatal)
Phenytoin (anticonvulsant)
Agranulocytosis, cutaneous ADRs
Procainamide (antiarrhythmic)
Hepatitis, agranulocytosis
Sulfamethoxazole (antibacterial)
Agranulocytosis, aplastic anaemia
Terbinafine (antifungal)
Hepatitis, cutaneous ADRs
Ticlopidine (antithrombotic)
Agranulocytosis, aplastic anaemia
Tolcapone (antiparkinsons)
Hepatitis (fatal),severe restriction in use
Trazodone (antidepressant)
Hepatitis
Trimethoprim (antibacterial)
Agranulocytosis, aplastic anaemia,
cutaneous ADRs
Thalidomide (immunomodulator)
Teratogenicity
Valproic acid (anticonvulsant)
Hepatitis (fatal), teratogenicity
For many drugs associated with IADRs, circumstantial evidence suggests a link with RM formation
Structure-toxicity relationships – evident and present a compelling case against RM positives
10
Structure-Toxicity Relationships – Example 1
Enol-carboxamide-containing NSAIDs
OH
O
N
S
Sudoxicam
Hepatotoxic (acute liver failure)
Withdrawn from Phase III trials
N
S
N
H
CH3
O O
OH
S
O
N
Meloxicam
N
N
H
S
CH3
“Clean” drug
CH3
O O
OH
S
N
Piroxicam
O
N
H
CH3
N
“Clean” drug
O O
11
Rationalizing the Differences in Toxicological Profile Through
Differences in Metabolism
OH
OH
S
N
O
N
N
H
OH
S
O
P450
CH3
N
S
O O
O
N
N
H
OH
OH
N
OH
S
CH3
N
S
O O
Sudoxicam
O
N
H
S
S
CH3
H
O O
O
H
Thioureas are toxic substances –
Can oxidize proteins, glutathione, etc
S
N
O
N
N
H
CH3
O O
OH
S
NH2
N
H
CH3
O O
Thiourea
Epoxide
OH
N
O
O
O
OH
N
CH3
P450
S
N
N
H
S
CH3
O O
Piroxicam
Principal metabolism in humans is hydroxylation on methyl
Very minimal thiazole ring opening
Obach, R. S., Kalgutkar, A. S., Ryder T. and Walker, G. W. “In Vitro Metabolism and Covalent Binding of Enol-Carboxamide Derivatives and
Anti-inflammatory Agents Sudoxicam and Meloxicam: Insights into the Hepatotoxicity of Sudoxicam.” Chem. Res. Toxicol. 2008, 21, 1890-1899.
12
S
Structure-Toxicity Relationships – Example 2
Antipsychotic agents
CH3
Clozapine
Agranulocytosis /Hepatotoxicity
N
N
N
Cl
(Black box warning – requires
intensive monitoring)
N
H
Quetiapine (Seroquel®)
N
O
N
Liu ZC, Uetrecht JP (1995) Clozapine is oxidized by activated human neutrophils to a reactive nitrenium ion that irreversibly binds to the
cells. J. Pharmacol. Exp. Ther. 275:1476-1483.
Gardner I, Leeder JS, Chin T, Zahid N, Uetrecht JP (1995) A comparison of the covalent binding of clozapine and olanzapine to human
neutrophils in vitro and in vivo. Mol. Pharmacol. 53:999-1008.
Commericial blockbuster
N
OH
S
N
N
O
CH3
Loxapine
N
“Clean” drug
Cl
13
Rationalizing the Differences in Toxicological Profile
Through Differences in Metabolism
CH3
Uetrecht J, Zahid N, Tehim A, Fu JM, Rakhit S. (1997) Structural features associated with reactive metabolite formation in clozapine analogues. Chem. Biol. Interact. 104:117-129.
CH3
CH3
N
N
Cl
N
N
N
N
P450
or
Peroxidase
Cl
N
N
Cl
N
H
N
H
GSH
S
O
N
Clozapine
N
H2N
N
H
COOH
Electrophilic iminium species
H
N
COOH
O
Bioactivation of clozapine catalyzed by peroxidases in neutrophils
Reactive metabolite responsible for covalent binding to neutrophils
N
N
O
N
N
OH
S
Quetiapine
N
CH3
N
O
Cl
Loxapine
Quetiapine and loxapine cannot form electrophilic iminium like clozapine does
14
RM Detection – Electrophile Trapping
• Reactive metabolites (with the exception of acyl glucuronides) are
unstable
– Need derivatization techniques for indirect characterization
• RM trapping with exogenous nucleophiles
– Can be used with diverse metabolism vectors
• Liver microsomes, S-9, hepatocytes, etc
• Glutathione, , N-acetylcysteine (soft nucleophiles)
– Traps soft electrophiles (e.g., Michael acceptors — quinones)
• Methoxylamine, semicarbazide, cyanide (hard nucleophiles)
– Traps hard electrophiles (e.g., aldehydes, iminium ion)
– LC-MS/MS and/or NMR methodology for structure elucidation of
O
conjugate
HS
O
NH
N
H
COOH
CH3ONH2
H
N
H2N
H2N
COOH
Glutathione
CN
NH2
O
Amines
Cyanide
15
RM Detection – Covalent Binding
• Limited to availability of radiolabeled drug candidate
– May not be suitable in early discovery
• In Vitro covalent binding can be assessed with
diverse metabolism vectors
– Effect of competing/detoxicating drug metabolizing enzymes on
covalent binding can also be examined
• In vivo covalent binding can be assessed in
preclinical species
• Covalent binding data is quantitative
– No information on nature of proteins modified
16
Utility of RM Detection Tools in Drug Discovery Identifying the Metabolic Basis for Mutagenicity
N
N
N
O
HN
CP-809,101
Cl
Selective and potent 5-HT2C Agonist
• Excellent in vivo pharmacology for weight reduction
• Excellent predicted human pharmacokinetics
•Potential as an anti-obesity agent
• Mutagenic in Salmonella Ames assay
• Requires Ariclor Rat S-9/NADPH
• Suggests DNA-Reactive Metabolites Formed
• Compound dropped from development
No toxicophore / Structural alert present; clean in DEREK assessment
GOAL
Need to elucidate mutagenic mechanism(s) for design of follow-on candidates
___ Available tools: [14C]-CP-809,101, RM traps (GSH, CH3ONH2, etc)
Kalgutkar, A. S., Dalvie, D., Aubrecht, J., Smith, E., Coffing, S. et al. Genotoxicity of 2-(3-Chlorobenzyloxy)-6-piperazinyl)pyrazine. A Novel 5-HT2C Receptor Agonist for the Treatment of
Obesity: Role of Metabolic Activation. Drug Metab. Dispos. 2007, 35, 848-858.
17
NADPH-Dependent Covalent Binding to CalfThymus DNA by [14C]-CP-809101
TABLE 1
S-9/NADPH-dependent covalent binding of [14C]-CP-809,10 to calf-thymus DNA
*N
N
HN
*
N
Test I
Test II
Mean DPMa/20 g DNA
Mean DPMa/20 g DNA
Vehicle (DMSO)
35 (34, 35)
23 (24, 24, 22)
CP-809101 (0.5 M) - S9
51 (54, 47)
52 (48, 48, 59)
CP-809101 (0.5 M) + incomplete S-9
40 (39,41)
42 (42, 41, 43)
CP-809101 (0.5 M) + complete S-9
105 (104, 106)
124 (126, 121)
CP-809101 (5.0 M) + complete S-9
495 (490,501)
462 (472, 451)
O
Cl
Incubation
CP-809,101
- S-9, without metabolic activation; + incomplete S-9 (-NADPH); complete S-9 (+
NADPH) NADPH was used in test I and NADPH regenerating system was used in test II.
a
mean DPM represents average from two to three separate experiments.
NADPH-dependent covalent binding to DNA suggests
P450-mediated bioactivation to DNA-reactive metabolite(s)
18
Deciphering CP-809101 Bioactivation Pathways
Cl
Cl
Cl
HO
HO
GSH
O
O
H
N
O
Cl
N
NH
CH2
N
S
NH
COOH
O
NH2
NH
N
COOH
Quinone-methide
S-9 / NADPH
O
H
N
O
NH
N
S-9 / NADPH
O
N
N
N
NH2
O
OH
N
Aldehyde
O
N
N
N
OH
N
O
N
N
N
O
N
CN
N
O
N
S-9 / NADPH
N
NH2
CH3ONH2
O
N
CP-809,101
N
Cl
O
N
N
N
OCH3
N
N
OH
CN
N
Nitrone
Covalent binding to DNA significantly attenuated in the presence of CH3ONH2 and GSH
19
Rational Chemical Modifications to Circumvent
Mutagenicity
F

F
N
N
F
F
N
N
O
N
HN
N

O
HN
Cannot form quinone-methide
Metabolic soft spots
(Minimal ring opening)
Non-mutagenic
in Ames Assay
Primary
Pharmacology
Maintained
PK Attributes
Maintained
Kalgutkar, A. S., Bauman, J. N., McClure, K. F.; Aubrecht, J. Cortina, S. R. and Paralkar, J. “Biochemical Basis for Differences in Metabolism-Dependent Genotoxicity by Two DiazinylpiperazineBased 5-HT2C Receptor Agonists.” Bioorg. Med. Chem. Lett. 2009, 19, 1559-1563.
20
RM Trapping and/or Covalent Binding Studies in Drug Discovery
Identifying Intrinsically Electrophilic Compounds - Influencing Scaffold
Design
N
N
N
O
CH3
O
R
CN
(1)
SAR Studies in a early discovery program
Compound 1 identified as meeting desired criteria for primary in vitro pharmacology
and progressed for further profiling (e.g., in vitro ADME, metabolism studies, etc) as
part of lead optimization efforts
21
Glutathione Trapping Studies on 1
100
M3-1
(A)
18.36 M2-1
19.81
50
0
100
(B)
Intensity
0
100
M4-1
1
HLM + NADPH + GSH
19.82
1
HLM – NADPH + GSH
1
GSH in buffer
M1-1
21.75
(C)
M4-1
24.02
(D)
M4-1
15.98
(E)
16.05
M4-1
50
0
100
HLM + NADPH
23.99
50
0
100
M3-1 M2-1
15.98
50
0
100
1
23.92
18.36
15.26
23.96
21.76
15.94
M5-1
50
M1-1
1
Cytosol + GSH
24.07
(F)
15.95
M4-1
50
1
0
23.95
24
14
16
18
Kalgutkar AS, Sharma R, Walker GS et al. Unpublished data
20
22
Time (min)
GST + GSH
26
22
Mass Spectra of M4-1 and M5-1
COOH
H2N
100
439.1021
M4-1
O
310.0600
N
HN
80
N
S
O
NH
O
R
CN
COOH
60
423.1437
298.0600
Intensity
40
212.1276
228.0434
257.0761
20
0 150
100
M4-1 (Exact mass + H+ = 638.2238)
200
250
509.1800 563.1909
594.2327
489.1540 552.1862
469.1495
638.2099
364.0707
300
350
400
450
500
550
600
COOH
M5-1
(- H2O)
H2N
O
80
650
636.2067
HN
N
S
O
60
N
O
R
OH
CN
NH
423.1436
COOH
M5-1 (Exact mass + H+ = 654.2182)
40
439.1019
298.0599
20
0150
281.0338
228.1229
193.8677
200
250
525.1748
340.8997
300
552.1854
350
400
m/z
579.1861
489.1560
396.3724
450
500
550
600
650
23
Additional Confirmation of Adduct Structure using
NMR
COSY
HMBC
Position
(1H) ppma
Group
 (13C) ppma
a
C? O
170.4
b
CH
3.30
52.9
c
CH2
1.88
26.8
d
CH2
2.34
31.5
e
C? O
171.8
f
NH
8.71
g
CH
4.61
51.6
h
C? O
169.1
i
NH
8.09
j
CH2
3.52
42.5
k
C? O
171.4
l
CH2
3.40,3.92
31.3
2
CH
8.85
159.8
4
C
167.3
5
C
91.5
6
C
173.4
7
CN
112.4
8
CH
5.52
74.3
9
CH
2.04
32.4
10 & 12
CH2
3.22,3.33,4.05, 4.13
42.0
13
CH
2.04
32.4
14 & 15
CH2
3.71,3.97
70.4
16
C? O
154.4
17
CH
4.78
67.3
18 & 19
CH2
1.18
21.4
a
Proton chemical shifts were measured relative to DMSO-d6
signal at 2.50 ppm. b Carbon chemical shifts were obtained from
the g-HSQC and g-HMBC spectra and measured relative to the
DMSO-d6 signal at 39.5 ppm.
2
NH2
HO
a
b
d
e
H
N
N
g
O h
HN i
O
O
j
k
OH
4
S
f
c
N
l
O
6
5
O
R
7C
N
24
The Cyanide Group in 1 is Essential for Nucleophilic
Displacement by GSH
COOH
CH3
N
N
O
CH3
GSH
N
N
CH3
N
O
N
GS
R
C
N
NH2
O
N
O
N
R
N
GS
C
N
HN
C
N
N
OH
HN
N
O
CH3
GSH
NH2
O
R
CH3
NO REACTION
HN
HN
O
R
C
N
COOH
COOH
N
N
S
CH3
Meisenheimer Complex
N
N
R
N
N
S
O
R
CH3
COOH
Cyanide substituent required for nucleophilic displacement by glutathione
Conclusions –
GSH adduct formation does not require “bioactivation”
4-Aryloxy-5-cyanopyrimidines can function as potential affinity labels (protein alkylation) or
cause GSH depletion
Cyano replacements in the current scaffold avoided
25
Eliminating Toxicity Risks in Drug Discovery Setting
• Structure-toxicity analyses teaches us that avoiding RM formation
with drug candidates represents one potential solution to
preventing drug toxicity
• Avoid chemical functionalities known to be
susceptible to reactive metabolites
– Tall order but avoids risk 
26
Examples of Functional Groups Susceptible to RM Formation
Anilines (masked anilines)
p-Aminophenols
Nitrobenzenes
Hydrazines (phenylhydrazines)
Benzylamines
Catechols
Cyclopropylamines
1,2,3,6-Tetrahydopyridines
2-Halopyridines and pyrimidines
Haloalkanes
Unsubstituted alkenes
Acetylenes
Imides
Formamides
Sulfonylureas
Thioureas
Methylenedioxy groups
Reduced aromatic thiols
5-Hydroxy(or methoxy) indoles
3-Methylindoles
Unsubstitued furans
Unsubstitued thiophenes
Unsubstitued thiazoles
Unsubstitued oxazoles
Thiazolidinediones
Fatty acids (medium to long chain)
Carboxylic acids
Hydroxylamines
Hydroxamic acids
Michael Acceptors
F
Hydroquinones
Bromobenzene
BENZENE !!!!!
N
H
N
OH
OH
COOH
O
R
R = H, OH
Lipitor®
Kalgutkar AS, Gardner I, Obach RS et al. (2005) Curr Drug Metab, 6, 161-225.
27
And What about the False Negatives?
H2N
I really don’t see a “ugly” looking structure here
O
O
O
O
NH2
I checked for glutathione conjugates in HLM and human
hepatocytes and saw none
I assessed covalent binding to HLM and human hepatocytes
and saw nothing of significance
Leone AM, Kao LM, McMillian MK et al. Evaluation of Felbamate and Other Antiepileptic Drug Toxicity Potential Based on Hepatic Protein Covalent Binding and Gene Expression.
Chem. Res. Toxicol. 2007, 20:600-608.
Obach, R.S., Kalgutkar, A. S., Soglia, J. R. and Zhao, S. X. “Can In Vitro Metabolism-Dependent Covalent Binding Data in Liver Microsomes Distinguish Hepatotoxic from Non-hepatotoxic Drugs?
An Analysis of Eighteen Drugs with Consideration of Intrinsic Clearance and Daily Dose.” Chem. Res. Toxicol. 2008, 21, 1814-1822.
Bauman, J., Kelly, J., Tripathy, S., Zhao, S., Lam, W., Kalgutkar, A. S. and Obach, R. S. “Can In Vitro Metabolism-Dependent Covalent Binding Data Distinguish Hepatotoxic from Non-Hepatotoxic
Drugs? An Analysis Using Human Hepatocytes and Liver S-9 Fraction.” Chem. Res. Toxicol. 2009, 22, 332-340.
But this is the anti-convulsant felbamate (Daily Dose > 3000 mg)
Within a year of its release in 1993
• 34 cases of aplastic anemia resulting in 13 deaths (Incidence rate 1:4800 – 1:37000)
• 23 cases of hepatotoxicity resulting in 5 deaths (Incidence rate 1:18000 – 1:25000
Black box warning (severe restriction in use)
• ~ 12,000 patients estimated to be on drug
28
In Vivo Observations on Felbamate
Conversion to RMs in Humans
metabolites in human urine
R
A heavy duty electrophile
O
O
S
HOOC
GSH
SG
NHCOCH3
2-Phenylpropenal
R = CH2OH or CO2H
O
CO2, NH3
NH2
O
HO
O
Amidase
O
NH2
O
CO2, NH3
O
NH2
O
ALD
O
NH2
O
Felbamate
Thompson CD, Barthen MT, Hopper DW, Miller TA, Quigg M, Hudspeth C, Montouris G et al. (1999) Quantification of patient urine samples of felbamate and three metabolites: acid carbamate and two
29
mercapturic acids. Epilepsia 40:769-776.
Diekhaus CM, Thompson CD, Roller SG, Macdonald TL (2002) Mechanisms of idiosyncratic drug reactions: the case of felbamate. Chem. Biol. Interact. 142:99-117.
And What about the False Negatives?
H2N
H2N
NOH
O
HN
O
O
HN
N
Ximelagatran
NOH
H2N
O
Et
O
H2O
NH
O
OH
HN
O
O
HN
OH
e
HN
N
N-Hydroxymelagatran
O
HN
N
O
Melagatran
No obvious structural alert
No evidence of glutathione conjugate formation in vitro or in vivo
Primary metabolic pathways in humans – ester hydrolysis, reduction to active drug “Melagatran”
Testa L, Bhindi R, Agostoni P, Abbate A, Zoccai GG, van Gaal WJ. The direct thrombin inhibitor ximelagatran/melagatran: a systemic review on clinical applications and an
evidence based assessment of risk benefit profile. Expert Opinion in Drug Safety 2007;6:397-406.
Ximelagatran (Exanta®), the first orally active thrombin inhibitor (anticoagulant) was
withdrawn due to several cases of hepatotoxicity
• Daily dose 20 – 60 mg BID
• Short term use (< 12 days) in humans did not indicate hepatotoxic potential
• Long term use (> 35 days) in human showed elevated hepatic enzyme levels in 0.5% of patients
• Withdrawal triggered from severe liver damage in a patient
• Immune component demonstrated upon pharmacogenomic analysis
30
Furthermore, How do we Handle the False Positives?
Adding insult to injury, some are commercial blockbusters
F
RM required for efficacy !!
O
O
H
O
O
OH
O
S
N
O
N
H
Paroxetine (Paxil®)
RM = Catechol /quinone
S
Cl
O
N
Clopidogrel (Plavix®)
Raloxifene (Evista®)
RM = Thiophene ring opening
O
Cl
N
Cl
OH
RM = quinone
CH3
N
O
N
N
N
N
N
N
N
N
H
S
CH3
Olanzapine (Zyprexa®)
RM = iminium
O
H2N
O
Prazosin (Minipress®)
RM = Furan Ring Opening
GSH conjugate/covalent binding demonstrated for all compounds
O
H
N
O
Aripiprazole (Abilify®)
RM = quinone imine 31
RM Detoxication as a Mitigating Factor for IADRs
The case of paroxetine
Detoxication
Detoxication
HO
H3CO
HO
HO
H3CO
O
O
HO
HO
O
O
O
HO
N
H
N
O
HO
Covalent Binding to
Hepatic Tissue
CYP2D6
O
O
S
GSH
COMT
O
O
GS
GS
F
HN
SG
+
HO
COOH
O
O
O
The case of Raloxifene
Bioactivation
O
O
O GSH
P4503A4
O
O
O
HO
S
Raloxifene
OH
S
HO
SG
N
OH
S
HO2C
HO
HO
O
O
O
O
OH
+
UGT
HO
S
O
Glu
S
Detoxication
32
Dose Size as a Mitigating Factor for IADR Potential of
New Drug Candidates
There are many examples of two structurally related drugs that possess a common structural alert prone to
bioactivation, but the one administered at the lower dose is much safer than the one given at a higher dose
Atypical anti-schizophrenia agents
CH3
N
N
Cl
N
N
H
Agranulocytosis in 2% of patients
Daily Dose = 300 mg
Olanzapine
CH3
N
N
N
N
H
Clozapine
S
CH3
Safe and Successful Drug
Sales > US $ 2 billion
Forms GSH conjugates via the iminium ion in a
manner similar to clozapine
Covalent binding to proteins
Only 3 cases of agranulocytosis
Higher than recommended dose
Daily Dose = 10 mg
33
Bioactivation Data Needs to be Placed in Proper Context —
Risk/Benefit Assessments (Qualifying Considerations)
• Nature of the medical need
– Life-threatening disease / unmet medical need
– First in class
• Target population
– Underlying disease state (immune-compromised patients)
• Is the drug candidate intended to provide proof of a novel
mechanism?
• What % of clearance mechanism involves bioactivation
– Existence of detoxication pathways; renal excretion, etc
• What is the daily dose of the drug?
– IADRs are rare for drugs dosed below 20 mg QD
34
For an account of a discovery strategy for dealing with RM
positive compound(s) as a drug candidate, please see:
Kalgutkar, A. S., Griffith, D. A., Ryder, T., Sun, H., Miao, Z., Bauman, J. N., Didiuk, M. T., Frederick, K. S., Zhao, S. X., Prakash, C., Soglia, J. R., et al.
Discovery Tactics to Mitigate Toxicity Risks Due to Reactive Metabolite Formation with 2-(2-Hydroxyaryl)-5-(trifluoromethyl)pyrido[4,3-d]pyrimidin-4(3H)-one
Derivatives, Potent Calcium-Sensing Receptor Antagonists and Clinical Candidate(s) for the Treatment of Osteoporosis.” Chem. Res. Toxicol. 2010, In Press.
F
F
CF3 O
Me
CF3 O
Me
CF3 O
Me
N
N
N
N
N
N
N
N
HO
N
F3C
HO
1
CaSR IC50 = 41 nM
HLM CLb = 12 mL/min/kg
RM positive
(GSH-EE conjugate peak area = 166810)
Predicted Human Dose = 45 mg QD
N
N
11
CaSR IC50 = 8 nM
HLM CLb = 7.4 mL/min/kg
13
CaSR IC50 = 64 nM
HLM CLb = 6.0 mL/min/kg
RM positive
RM negative
(GSH-EE conjugate peak area = 3000)
Predicted Human Dose = 10 mg QD
O
X
HO
X = CH, N
X
CYP
X
CYP
OH
NH
GSH
or
GSH-EE
X
S
O
HO
O
HO
COOR
O
NH
OH
NH2
HO2C
Reactive metabolite (RM) formation pathway
GSH: R = H
GSH-EE: R = CH2CH3
35