Transcript A Non-Discretionary Discipline in Tomorrow`s Drug

Pharmacogenomics in Action –
A Non-Discretionary Discipline
in Tomorrow’s Drug
Development
Lloyd Curtis
PRISM May 2009
Disclaimer
 All
views presented are personal and not
necessarily those of GSK
2
Pharmacogenomics - Definition
 There
is now an established regulatory
definition of Pharmacogenomics
 ICH HARMONISED TRIPARTITE
GUIDELINE – E15 (November 2007)

DEFINITIONS FOR GENOMIC
BIOMARKERS, PHARMACOGENOMICS,
PHARMACOGENETICS, GENOMIC DATA
AND SAMPLE CODING CATEGORIES
• http://www.pmda.go.jp/ich/e/step4_e15_e.pdf
3
ICH HARMONISED TRIPARTITE
GUIDELINE – E15
 Pharmacogenomics

(PGx) is defined as:
The study of variations of DNA and RNA
characteristics as related to drug response.
 Pharmacogenetics
(PGt) is a subset of
pharmacogenomics (PGx) and is defined
as:

The study of variations in DNA sequence as
related to drug response.
4
ICH HARMONISED TRIPARTITE
GUIDELINE – E15
 PGx

drug discovery, drug development, and
clinical practice.
 Drug

and PGt are applicable to:
response includes:
drug absorption and disposition (e.g.,
pharmacokinetics, (PK)), and drug effects
(e.g., pharmacodynamics (PD), drug efficacy
and adverse effects of drugs).
5
Pharmacogenomics – The Promise
 Improve
the diminishing productivity of
drug development
 Reduce the high proportion of patients
who:


Receive no benefit from administered drugs
Experience adverse reactions
6
Pharmacogenomics & Regulators
 USA 

FDA Critical Path Initiative – 2004
The Interdisciplinary Pharmacogenomics
Review Group
Pharmacogenomics Working Group
 Europe

The Pharmacogenomics Working Party
 Japan

- EMEA Road map – 2005
– PMDA
The Pharmacogenomics Discussion Group
7
FDA initiatives in
Pharmacogenomics







Workshops - use of biomarkers in drug
development and clinical practice
Issued guidances
Voluntary Genomic Data Submission Program
Developed online educational tools
‘Re-labelling’ as well as inclusion in new labels
Pilot process to qualify novel biomarkers
Creation of consortia - e.g International Serious
Adverse Event Consortium
8
Recognized challenges

Qualification of biomarkers

extent of information needed to understand clinical
utility

What clinical trial data are required to qualify
biomarkers
 Acceptable clinical trial design


prospective Randomised Controlled Trial,
observational cohort, retrospective?
Can modelling and clinical trial simulation be
used as evidence of biomarker qualification?
9
Safety example - Antiretroviral drug
abacavir





Abacavir hypersensitivity reaction (ABC HSR)
affected 5 to 8% of clinical trial subjects
Multi-organ clinical syndrome – typically fever
and/or rash and/or constitutional, GI and/or
respiratory symptoms
Rechallenge contraindicated and can be fatal
Clinical diagnosis imprecise due to concurrent
drugs/illnesses
Results in 2-3% false positive rate in blinded
clinical trials
10
Abacavir - background

Effective risk management programme created
(educational materials, labelling,
pharmacovigilance, etc)
 Pharmacogenetic research identified HLA-B*5701
allele more common in Caucasian patients with
clinically suspected ABC HSR (2001)
 Sensitivity/specificity of test varied between
studies and racial populations – limitations of
case-ascertainment (and unethical to
rechallenge)
11
ABC HSR – PREDICT-1study

Prospective Randomized Evaluation of DNA
Screening In a controlled Clinical Trial to
determine the clinical utility of HLA-B*5701
screening prior to ABC-containing therapy
 Two co-primary endpoints:
• Rate of clinically-suspected HSR
• Rate of immunologically confirmed HSR
(HSR plus positive patch test reaction)
12
PREDICT-1 study design
ABC-containing regimen
HSR monitoring according to
Standard of Care
ABC-naïve Blinded randomisation (1:1)
Subjects
ABC-containing regimen
N=1956
HSR monitoring according to
Standard of Care plus
PGx screening
6 weeks of study observation
Exclude subjects
with HLA-B*5701
Commence ABC in
HLA-B*5701 negative
Subjects
13
PREDICT-1 results
9
OR 0.40 (0.25, 0.62)
P < 0.0001
8
Control arm (Standard of Care)
7
Prospective HLA-B*5701
screening arm
Incidence (%)
6
5
OR 0.03 (0, 0.18)
P < 0.0001
4
3
2
7.8%
3.4%
(66/847) (27/803)
1
0
Clinically Suspected
HSR
2.7%
(23/842)
0.0%
(0/802)
Immunologically Confirmed
HSR
14
ABC HSR – SHAPE study
 Study
of Hypersensitivity to Abacavir
and Pharmacogenetic Evaluation:


Retrospective, case-control trial in selfreported Black and White patients to
assess the generalisability of HLAB*5701
Note: HLA-B*5701 carriage frequency
lower in Blacks than Whites
15
SHAPE study design
ABC skin
patch test
Identify ABC-tolerant
subjects who provided
PGx consent and sample
CASES
Black & White subjects
enrolled in KLEAN, ALOHA,
CNA30027, CNA30032
Skin patch
test positive
(IC-HSR)
PGx
evaluation
Skin patch
test negative
PGx
evaluation
CONTROLS
Black and White subjects
with clinically-suspected
ABC HSR (CS-HSR)
PGx
evaluation
Up to 200 controls
for each race
16
SHAPE pharmacogenetic analyses
SPT (+)
SPT (-)
CS-HSR
Controls
42/42 (100)
15/85 (18)
57/129 (44)
8/202 (4)
Sensitivity (95% CI)
1.0
(0.92-1.0)
-
0.44
(0.35-0.53)
N/A
Specificity (95% CI)
N/A
-
N/A
0.96
(0.92-0.98)
Odds Ratio (95% CI)
1945
(110-34352)
-
19
(8-48)
Ref
Black Subjects
HLA-B*5701 Present, n (%)
5/5 (100)
5/63 (8)
10/69 (14)
2/206 (<1)
Sensitivity (95% CI)
1.0
(0.48-1.0)
-
0.14
(0.07-0.25)
N/A
Specificity (95% CI)
N/A
-
N/A
0.99
(0.97-1.0)
Odds Ratio (95% CI)
900
(38-21045)
-
17
(4-164)
Ref
White Subjects
HLA-B*5701 Present, n (%)
17
EMEA and ABC HSR
EU Summary of Product Characteristics
update, Jan 2008

‘Before initiating treatment with abacavir,
screening for carriage of the HLA-B*5701
allele should be performed in any HIV-infected
patient, irrespective of racial origin’
(Note: in EU ‘should’ means mandated)
18
FDA and ABC HSR
 US
Prescribing Information update, July
2008

‘Prior to initiating therapy with abacavir,
screening for the HLA-B*5701 allele is
recommended; this approach has been found
to decrease the risk of a hypersensitivity
reaction’.
19
Advancing technology - Potential
for Routine Genome-Wide Analysis
 Current
genotyping technologies make it
possible to investigate adverse drug
reaction (ADR) genetics during the course
of clinical trials and post-approval
pharmacovigilance

Candidate gene panels
• Select up to 10,000’s genetic markers in 100’s of
candidate genes

Genome-wide panels
• 0.5-1+ million markers
21
The POPRES Initiative

Population Reference Sample
 Facilitate exploratory population, disease, and
pharmacogenetic research through access to



DNA from a variety of representative population
samples
Genotyped for genome-wide, as well as other focused
SNP panels
Genotypic and basic demographic data from
POPRES is publicly available via dbGaP
22
Abacavir HSR Study Design
 Sample

Cases
• 22 abacavir-treated, White, HIV+ subjects that
experienced HSR while on treatment

Reference sample (i.e. “Controls”)
• 203 POPRES subjects of European origin (US,
Canada, Australia)
 Genotype

data
Affymetrix 500K SNP chip set
23
“Standard” Genome-Wide Analysis

Analyze each marker of the 500K panel in all
ABC HSR cases versus population reference
controls
 Identify markers and regions most strongly
associated with ADR
 Methods



Filter out markers with significant deviations from
Hardy-Weinberg proportions: α = 0.05/5x105
Single SNP allelic and genotypic exact tests
Can we identify the known HLA-B association
with 22 cases and 203 population reference
controls?
24
WGS Identifies HLA-B Region
among Top Candidates
25
Flucloxacillin and Liver Injury

Flucloxacillin important cause of drug induced
liver injury (DILI) in Europe and Australia

8.5 cases/100,000

GWS and Candidate gene studies in parallel
 Cases – 51 definite/possible fluclox DILI
 GWS controls 282 matched gender and
population origin
 Candidate gene controls 64 patients exposed
fluclox without DILI
Daly et al, 6th Wellcome Trust Conference on Pharmacogenomics (CSH) 2008
26
GWS and Candidate Gene Studies Identify
HLA-B*5701 As the Major Risk Allele of DILI
Caused by Flucloxacillin
 GWS


Illumina Human1M chips
84% cases vs <5% controls, odds ratio 36
 Candidate


genes
major histocompatibility complex (MHC)
region and a few other immune-system
related genes
84% cases vs 6% controls
Daly et al, 6th Wellcome Trust Conference on Pharmacogenomics (CSH) 2008
27
Conclusion on Potential for Routine
Genome-Wide Analysis





Large-scale genotyping offers new opportunities to
consider the impact of PGx on ADRs
ADR genetic risk factors with reasonably large
effects can be identified with relatively small case
sample sizes
Availability of dense, genome-wide genotype data
on suitable population reference samples can
facilitate rapid, exploratory research
This strategy would have likely identified HLA-B
region for abacavir-associated HSR with 15-20
cases
Application of sequential methods as ADRs accrue
can lead to early identification of ADR PGx risk
factors
28
Pharmacogenomics Promise
Unfulfilled?
FDA Drug Safety Newsletter – Winter 2008
 Safety Examples
 Drug resistance mutations in HIV
 Rapid and slow metabolizers – codeine
 Warfarin VKORC1 (target) & P450 2C9
(metabolism)- recommended (2)
 HLA-B*1502 and carbamazepine associated
Stevens Johnson syndrome in populations of
specific Asian ancestry - recommended (2)

29
Newsletter Efficacy examples
 Imatinib
for bcr-abl tyrosine kinase in
several tumour types – information only (3)
 Cetuximab for epidermal growth factor
receptor (EGFR) in


head and neck cancer - information only (3 )
colorectal cancer– required (1)
 Trastuzumab
for variants in the Her2
receptor in breast cancer – required (1)
30
Conclusions
 Strong
regulatory pressure to pursue
pharmacogenomic endeavours in drug
development – particularly FDA
 Modest results so far with more hope for
safety than efficacy
 Pharmacogenomics will be ‘nondiscretionary’ driven by regulators rather
than demonstrable success
31
Conclusions

Support technologies developing at rapid pace viz:
 Make available a biomarker data repository to store
and retrieve flexibly pharmacogenomic data types
including gene chip data, sequencing data and other
“omics” data
 Enable the integration of traditional phenotypic clinical
trial data with biomarker data including genotypic data
 Provide flexible & powerful data visualisation, mining
and statistical tools to seek for and find correlations
between AEs and gene expression
 Ensure workflow tools exist to document the
provenance of an analysis and provide repeatability
32
Contributors
POPRES Project
 Sponsors



Project lead


Matt Nelson
Team members








Eric Lai
Dan Burns
Linda Briley
Clive Bowman
Meg Ehm
Kelley Johansson
Brendan Jones
Karen King
Heide Stirnadel
Additional support,
former team members







Donna Backshall
Devon Kelly
Michael Klotsman
Yuka Maruyama
Annie McNeill
Tony Morris
Jill Ratchford
Sequential PGx Methods



Clive Bowman
Silviu-Alin Bacanu
Michael Lawson
Abacavir Case Study









Cindy Brothers
Charles Cox
Kirstie Davies
Seth Hetherington
Jaime Hernandez
Arlene Hughes
Mike Mosteller
Bill Spreen
Liling Warren
Special Acknowledgement
Thanks to the clinical trial
and reference sample
participants who provided
both informed consent and
blood samples for the PGx
research
Thanks also to the clinical
investigators and their
study staff, the GSK
clinical development teams
and all POPRES
collaborators
GSK PGx




Allen Roses
Li Li
Stephanie Chissoe
David Yarnall
33
The chemical name of abacavir sulfate is (1S,cis)-4-[2amino-6-(cyclopropylamino)-9H-purin-9-yl]-2cyclopentene-1-methanol sulfate (salt) (2:1). Abacavir
sulfate is the enantiomer with 1S, 4R absolute
configuration on the cyclopentene ring. It has a molecular
formula of (C14H18N6O)2•H2SO4 and a molecular
weight of 670.76 daltons. It has the following structural
formula:
34