Transcript Dr Richard Fitzgerald

Same Medicine, Different Result
Pharmacogenetics: Where Are We Now?
Dr Richard FitzGerald
Molecular & Clinical Pharmacology
Institute of Translational Medicine
University of Liverpool
[email protected]
The drugs don’t work.......
....... they just make it worse.
The problem: variability

‘If it were not for the great
variability among
individuals, medicine might
as well be a science and not
an art.’
Sir William Osler, 1892
Pythagoras (6th Century B.C.)



“…..be far from fava beans
consumptions”
Met death in Ancient Italy
because he refused to cross a
field of beans
Many theories:
 Contained souls
 Looked like testicles
 flatulence
 Medical reason
FAVISM
Fava beans
RBC
haemolysis
‘Chemical Individuality’
First suggested by Sir
Archibald Garrod that
genetics may affect
chemical
transformations
 He used the example of
alkaptonuria (1902)

‘One gene, one enzyme’
Modern pharmacogenetics
Types of Genetic Variation
Drug Response: a complex trait?
The early years: one gene, one disease

Robert Smith investigated debrisoquine (a commercially available
anti-hypertensive)

He took the tablet, along with most of his laboratory staff

He collapsed and became markedly hypotensive. Nobody else did.
CYP2D6 Major Alleles
Nortriptyline pharmacogenetics
Codeine phosphate
Drug metabolising enzymes
Most DME have clinically relevant polymorphisms
Those with changes in drug effects are separated from pie.
Xanthine
oxidase
Thiouric
acid
Azathioprine
TPMT
6-Mercaptopurine
6-Me MP
HGPRT
6-thioinosine nucleotide
IMPDH
6-thioguanine
nucleotides
Immunosupression
Clinical benefit
TPMT (Thiopurine methyltransferase)
Allelic polymorphism
Low
TPMT
1/300
Severe Bone
Marrow
Suppression
+
Intermediate
TPMT
11%
High risk
of marrow
suppression
High
TPMT
89%
?very high
TPMT
Low risk
Low risk
? poor
responders
clinical response
-
PGx: current applications
Abacavir Hypersensitivity





Nucleoside analogue
Reverse transcriptase
inhibitor
Hypersensitivity 5%
Fever, skin rash, gastrointestinal symptoms,
eosinophilia within 6
weeks
Re-challenge results in a
more serious reaction
Abacavir Hypersensitivity
Clinical genotype
NH
N
N
H2N
N
Association with
HLA-B*5701
N
CH2OH
Clinical phenotype
Causal chemical
Incidence before and after testing for HLA-B*5701
Country
Pre testing
Post testing
Reference
Australia
7%
<1%
Rauch et al, 2006
France
12%
0%
Zucman et al, 2007
UK (London)
7.8%
2%
Waters et al, 2007
PGx: effects on drug usage
8000
350
7000
300
6000
250
5000
200
Combivir
150
Kivexa
Truvada
HLA*
4000
3000
100
2000
50
1000
0
J
F
M
A
M
J
J
A
S
O
N
D
J
F
M
A
M
J
J
A
S
O
N
D
J
F
M
A
M
J
J
A
S
O
N
D
0
2005
2006
2007
Data from RLBUHT courtesy of Prof Saye Khoo
PREDICT-1
Abacavir Genetics: Why so Rapidly Implemented?
Implemented even before RCT evidence
 In some cases, observational study designs may provide adequate
evidence
Successful implementation was because of several factors:
 Good and replicated evidence of a large genetic effect size
 Clinician community amenable to rapid change in clinical practice
 Vocal and knowledgeable patient lobby
Carbamazepine-induced hypersensitivity reactions
5% of patients on carbamazepine (CBZ) develop hypersensitivity reactions
10% in prospective SANAD study (UK)

Clinical manifestations
Maculopapular exanthema
usually mild

Hypersensitivity reaction (HSS)
1/1000 patients
Fever, hepatitis, eosinophilia

Stevens-Johnson syndrome
Toxic epidermal necrolysis
5-30% fatality rate
FDA warning

PATIENTS WITH ASIAN ANCESTRY SHOULD BE SCREENED FOR THE
PRESENCE OF HLA-B*1502 PRIOR TO INITIATING TREATMENT WITH
Carbamazepine.
To prospectively identify subjects at risk for SJS
4877 CBZ naive subjects from 23 hospitals
The Taiwan SJS Consortium
HLA-B*1502 testing → 0 incidence of SJS/TEN
University of
Liverpool (SANAD,
EUDRAGENE, Swiss,
WT Sanger, Harvard)
 EPIGEN Consortium
(Ireland, Duke
University, UCL,
Belgium)

Faculty of 1000 -top 2% of published articles in biology and medicine
 American Academy of Neurology meeting- voted as one of the top
articles in neurology this year

HLA-A*3101
22 patients with HSS
2691 healthy control subjects
HLA-A*3101
43 patients with MPE
1296 healthy control subjects
McCormack et al. NEJM 2011
Pooled analysis of case-control studies
P=
P=0.03
P=8 x10-7
P=8 x10-5
P=1x10-7
McCormack et al. NEJM 2011
GWAS identifies HLA-A*3101 allele as a genetic risk factor for CBZinduced cutaneous adverse drug reactions in Japanese population
HLA-A*3101
Ozeki et al. Hum Mol Genet 2011
Conclusions






HLA-A*3101 - a prospective marker for CBZ
hypersensitivity
Associated with several phenotypes
Further work needed to enable clinical use
Need for consortia
Possibility of rare variants and CNVs
(exome-sequencing/WGS)
Mechanistic studies to follow genetics
Flucloxacillin-Induced Cholestatic Hepatitis:
Whole Genome Scan
Daly at al, 2009
Illumina 1 million SNP array
Strong (P=10-30) association with SNP in LD with HLA-B*5701
Weaker association with novel marker on chromosome 3 (p <
1.4 x 10-8 )
Weak association with copy number polymorphism
Performed in collaboration with the Serious Adverse Event Consortium
1. Implicated SNP is in the SLCO1B1 gene (transporter)
2. Shown with simvastatin 40mg and 80mg
3. C variant may account for 60% of the cases of myopathy
Clopidogrel Pharmacogenetics
Stent Thromb HR 2.61; 95% CI 1.61-4.37, P<0.00001
All events: HR 1.57; 95% CI 1.13-2.16, P=0.006
Conclusions

Clear adverse effect of the CYP2C19*2 polymorphism on clinical and
pharmacodynamic outcomes
 PD Meta-analysis limited by multiple outcome measures
Potential utility in CYP2C19*2 as marker of clopidogrel non-response
and risk of adverse outcome
 Translation into clinical practice

 Increase dose of clopidogrel from 75mg/day to 150mg/day
– Evidence from CURRENT-OASIS 7 trial
– Bleeding risk
 Use of alternative anti-platelet drugs (Prasugrel, Ticagrelor)
–
–
–
–
Better platelet inhibition
Higher rates of bleeding (+ other adverse effects)
Benefit may be only seen in those with the CYP2C19*2 allele
Cost
Warfarin: a more complex variation

Widely used drug

A variety of acute/chronic indications

Large numbers of patients

6% of all patients over 80 years of age

Narrow therapeutic index

Drug interactions and alcohol

Efficacy
• Bleeding complications:
10-24 per 100-patient years
• 10% of all ADR-related
hospital admissions
The clinical phenotype

10-50 fold variability in dose requirements

Increased age; decreased requirements
 8% decrease in warfarin dose per decade
 Enhanced responsiveness (PD)
 Reduced clearance (PK)
Warfarin and metabolism by
CYP2C9
CYP2C9*1
Wild Type Arg144 Ile359
CYP2C9*2
Arg144
Cys
: interaction with cytochrome
P450 reductase
CYP2C9*3
Ile359
Leu
: substrate binding site
: affects Km, Vmax
Variant alleles have 5-12% of the activity of wild-type
Steward et al, Pharmacogenetics (1997), 7, 361-367
Warfarin and pharmacokinetics
CYP2C9
genotype
Number of
patients
Aggregate mean
dose (mg)
CYP2C9*1*1
639
5.5
CYP2C9*1*2
207
4.5
CYP2C9*1*3
109
3.4
CYP2C9*2*2
7
3.6
CYP2C9*2*3
11
2.7
CYP2C9*3*3
5
1.6
Warfarin and pharmacodynamics
Polymorphisms in vitamin K epoxide reductase
(VKOR)C1
 Associated reductions in warfarin dose
 Accounts for greater variance in dose than CYP2C9
 Variation in genes encoding γ-glutamyl carboxylase
and factors II, VII and X

Genetic and Environmental Factors and
Dose Requirements of Warfarin
mg/week
VKORC1 SNP rs 2359612 vs. warfarin dose
50
45
40
35
30
25
20
15
10
5
0
AA
(n=29)
AG
GG
(n=96)
(n=75)
Independent effects of VKORC1 and CYP2C9:
VKORC1:
p<0.0001, r2 = 0.29
CYP2C9:
p=0.0003, r2 = 0.11
Age:
Body weight:
r2
p<0.0001, = 0.10
p=0.0018, r2 = 0.05
55%
Wadelius et al. 2005
Warfarin: multiple genes/factors
GENETIC
 Cytochrome P450
polymorphisms
 Vitamin K epoxide
reductase
 Phase II metabolising
genes
 Drug transporters
 Clotting factors
 Disease genes
ENVIRONMENTAL
 Sex
 Age
 Smoking
 Interacting drugs
 Alcohol
 Compliance
 Diet
Test interpretation
The potential for complication
Will pharmacogenetic testing be
any better than more intensive
INR monitoring?
Pharmacogenetic algorithm was superior to clinical algorithm or
fixed dosing
Greatest benefit seen in 46% of the population who require either
<3mg/day or >7mg/day
Two Randomised Controlled Trials
COAG




NIH-sponsored US trial
1200 patients
Genetic algorithm vs clinical
algorithm
%TIR as primary outcome
measure
EU-PACT





EU FP7 sponsored EU trials
3 trials: warfarin,
phenprocoumon,
acenocoumarol
900 patients in each (2700
total)
Final study design
completed
%TIR as primary outcome
measure
Closing The Loop
Show an
association
Replicate the
association
Demonstrate
clinical
validity and
utility
Identify a
variant
Demonstrate
a positive
clinical
outcome
Pre-clinical
Phases I, II, III
Phase IV
New technologies:
Pharmacogenomics
Proteomics
Metabolomics
Systems Biology
Minimise risk and maximize benefit
Uncertainty reduced but not abolished
Advances in Technologies
14 billion bases/day
PGx and Prospective Utility







Drug development process
Potential prospective use of PGx to enhance success
Increase confidence
US$1 billion to market a new drug
Target discovery
Proof of concept
Candidate gene/whole genome association
Current Status of Genetic Tests
“Today, there is no mechanism
to ensure that genetic tests are
supported by adequate evidence
before they are marketed or that
marketing claims for such tests
are truthful and not misleading.
Misleading claims about tests
may lead health-care providers
and patients to make
inappropriate decisions about
whether to test or how to
interpret test results.”
Science, 4 April 2008
Personalised Medicines: The Future?
Many recent advances
Here to stay, and likely to be supported by increasing evidence
Evolutionary process, not revolutionary
Lot of cynicism about personalised medicine approaches
Evidence being required is much greater with other tests
Personalised vs. Empirical Paradigms
Empirical (intuitive) medicine
Observation
Trial and
error
response
Action
Personalised (precision) medicine
Observation
Test
Action
Predictable
response
Terminology
Personalised Medicine
not
Personal Medicine
• We cannot truly personalise medicines
• No test or prediction rule will be 100%
effective
Arno Motulsky
“ What we know about the
genome today is not enough
for all the miracles many
expect from this field. There’s
a lot about what regulates
the genes and how they
interact that we still need to
understand. We won’t have
the answers by tomorrow.”
29th April 2008