Clinical Pharmacogenetics

1
Clinical Pharmacogenetics

• David A Flockhart MD, PhD
• Chief, Division of Clinical Pharmacology
• Professor of Medicine, Genetics and Pharmacology
• Indiana University School of Medicine

2
Pharmacogenetics, 1990
3
Pharmacogenomic Journals, 2006
4
SNP Variability in The Human Genome December 2005

• 2.85 billion base pairs
• 22,000 genes
• 1.7 of the genome codes for protein
• 3.3 of the genome is as conserved as the 1.7
  that codes for protein
• On average 1 SNP/1.2kb
• 10 - 15 million SNPs that occur at gt 1 frequency
• 450,000 SNPs in MCS (Multiply Conserved Regions)

5
SNP Variability In Exons

• 150,000 SNPs in known exons
• 48,451 non-synonymous SNPs
• 1113 introduce a stop codon
• 104 disrupt an existing STOP

6
Hierarchy of Pharmacogenetic Information from
Single Nucleotide Polymorphisms (SNPs)
7
Multiple Copies of human DNA can result in
significant inter-individual variability

• 3,654 autosomal segmental regions,
• 800 at a frequency of at least 3.
• In 95 individuals analyzed, the two most diverse
  genomes differed by at least 9 Mb in size or
  varied by at least 266 loci in content.
• 14 copied regions harbor 21 of the known human
  microRNAs, raising the possibility of the
  contribution of microRNAs to phenotypic diversity
  in humans.

Am J Hum Genet. 2007 Jan80(1)91-104.
i.e. We are NOT 99 the Same
8
Why Pharmacogenomics?

• To divine mechanism
• To predict therapeutic response

9
Pharmacogenetic Principle 1 Value Decreases when
Current Predictive Ability is High
Cancer Chemotherapy
Clinical Value of a Pharmacogenetic Test
Antidepressants/5HTR
Azathioprine/TPMT
ß-blockade/ß Receptor
Current Clinical Ability to Predict Response
Meyer UA and Flockhart DA, 2005
10
Pharmacogenomic Methods

• Candidate gene approach
• Pathway (multiple candidate gene) approach
• Whole genome approach
• Combinations of the above

11
Drug Response Pathways can Involve many Candidate
Genes
www.pharmgkb.org
12
Methods in Pharmacogenetics

• Population distribution analysis with Normit
  plots using a valid probe to detect phenotypic
  polymorphism (gt 1 of population)
• Identification of gene and variants
• Family and twin studies to confirm genetic
  characteristics (dominant, recessive, Mendelian,
  maternal etc.)
• Development of a genetic test for DNA variants
• Correlation between genotype and phenotype
• Application in Clinical Practice

13
Polymorphic Distribution
Antimode
14
The Value of Normit Distribution Plots
Population Distribution of CYP2C19 phenotype
Flockhart et al Clin Pharmacol Ther
199557662-669
15
Skewed Distribution
16
Example 1 of a Skewed Distribution Heterogeneity
in response to Inhaled Corticosteroids
Weiss ST et al. Hum Molec Genetics 2004
131353-1359
17
Using the extremes of a phenotypic distribution
as a strategy to identify pharmacogenetic
predictorsExample Iressa and the EGF receptor
18
EGFR mutations in tumors that are sensitive or
insensitive to kinase inhibitors.
All activating variants are in exon 9 and code
for 4 specific aa changes
Minna JD et al. Science 20043041458-1461
19
EGF Receptor Variant predicts Iressa Response
The oncogene addiction model
20
Lessons

• Germline genetic variation matters, but so do
  somatic mutations in tumors
• Extremes of phenotype are often viewed as
  discardable data, but outliers should be viewed
  as important research stimuli
• Patients who are outliers matter

21
Genetics and Drug Absorption
22
Digoxin Transport across the GI lumen
P-gp Transport
ATP
?
ADP
Passive Diffusion
Enterocyte
GI Lumen
23
P-Glycoprotein Pharmacogenetics Effect of a
wobble (no coding change) SNP in exon 26
Fig. 3. Correlation of the exon 26 SNP with
MDR-1 expression. The MDR-phenotype (expression
and activity) of 21 volunteers and patients was
determined by Western blot analyses. The box plot
shows the distribution of MDR-1 expression
clustered according to the MDR-1 genotype at the
relevant exon 26 SNP. The genotype-phenotype
correlation has a significance of P 0.056 (n
21).
Eichelbaum et al. Proc Nat Acad Sci March, 2000.
24
0.25 mg of digoxin po at steady state
Eichelbaum et al, Proc Nat Acad Sci, 2000March
25
Digoxin Transport across the Blood-Brain Barrier
P-gp Transport
ATP
?
ADP
Passive Diffusion
Brain
Blood
26
Genetics and Drug Elimination
27
Cytochrome P450 2D6

• Absent in 7 of Caucasians
• Hyperactive in up to 30 of East Africans
• Catalyzes primary metabolism of
• propafenone
• codeine
• ?-blockers
• tricyclic antidepressants
• Inhibited by
• fluoxetine
• haloperidol
• paroxetine
• quinidine

28
CYP2D6 Pharmacogenetics
29
New CYP2D6 variants continue to appear.
From Zanger et al Clin Pharmacol Ther. 2004
Aug76(2)128-38.
30
CYP2D6 Alleles

• 51 as of October, 2004
• 24 alleles have no activity
• 6 have decreased activity
• 1, 2, 4 and others have copy number
  polymorphisms
• The 2 variant can have 1,2,3,4,5 or 13 copies
  i.e increased activity

31
Oligonucleotide array for cytochrome P450
genotesting
From Flockhart DA and Webb DJ. Lancet End of
Year Review for Clinical Pharmacology, 1998.
32
From Dalen P, et al. Clin Pharmacol Ther
63444-452, 1998.
33
Paroxetine and CYP2D6 genotype change the plasma
concentrations of endoxifen
Stearns, Flockhart et al. J Natl Cancer Inst.
200395(23)1758-64.
34
Lessons from CYP Pharmacogenetics

• Multiple genetic tests of one gene may be needed
  to accurately predict phenotype
• Gene duplication in the germline exists
• The environment in the form of Drug Interactions
  can mimick a genetic change

35
Vitamin K Carboxylase and CYP2C9 Genotypes
altered Warfarin Dose
Rieder et al. N. Eng J. Med 2005352 2285-2293
36
Genetic alterations in Phase 2 enzymes with
clinical consequences UGT1A1 NAT-2SULT1A1COMT
TPMT
37
UGT1A1 TA repeat genotype alters irinotecan
neutropenia/activity
41.9
P0.045
45
40
33.8
35
30
Objective response ()
25
20
14.3
15
10
5
0
6/6
6/7
7/7
UGT1A1 genotype
N524
McLeod H. et al, 2003.
38
N-Acetylation PolymorphismNAT-2

• Late 1940s Peripheral Neuropathy noted in
  patients treated for tuberculosis.
• 1959 Genetic factors influencing isoniazid
  blood levels in humans. Trans Conf Chemother
  Tuberc 1959 8, 5256.

39
NAT-2 substrates(All have been used as probes)

• Caffeine
• Dapsone
• Hydralazine
• Isoniazid
• Procainamide

40
Incidence of the Slow Acetylator NAT-2 phenotype

• 50 among Caucasians
• 50 among Africans
• 20 among Egyptians
• 15 among Chinese
• 10 among Japanese

41
Onset of Positive ANA Syndrome with Procainamide.
Woosley RL, et al. N Engl J Med 2981157-1159,
1978.
42
Thiopurine Methyl Transferase

• Homozygous mutants are 0.2 of Caucasian
  Populations
• Heterozygotes are 10
• Homozygous wild type is 90
• Metabolism of Azathioprine
• 6-Mercaptopurine

43
Thiopurine Methyl Transferase Deficiency
From Weinshilboum et al. JPET222174-81. 1982
44
Germline TPMT genotype associates with incidence
of toxicity
45
Examples of Human Receptors shown to be
genetically polymorphic with possible alterations
in clinical phenotype

• G-proteins
• Angiotensin II receptor and angiotensinogen
• Angiotensin converting enzyme
• ?2receptor
• Dopamine D4 receptor
• Endothelial NO synthase
• 5HT4receptor

46
2SNPs 10 possible hapoltypes
Haplotypes
Diplotypes
Ying-Hong Wang PhD, Indiana University School of
Medicine
47
Observed b1AR Haplotypes in Caucasians and
African American Women (WISE study)
Terra et al. Clin. Pharmacol. Ther. 7170 (2002)
48
Of 10 theoretical diplotypes, only 4 were
present in the study population
Haplotypes
Diplotypes
Ying-Hong Wang PhD, Indiana University School of
Medicine
49
Diplotype predicts Beta-blocker effect
Johnson et al. Clin Pharmacol Ther.
2003,7444-52.
50
Lesson Diplotype may be a better predictor of
effect than GenotypeA SNP that tags a Haplotype
(tagSNP) may be an economical means of screening
51
ß2 receptor Gln27Glu (79CG) genotype predicts
mortality during ß-blockade after MI.
Lanfear DE et al. JAMA September 28th,
20052941526-1533.
52
Non-synonymous coding region polymorphisms in
long QT disease genes
SCN5A
I
N
a
KvLQT1
HERG
MiRP1
minK
I
K
r
I
K
s
Dan Roden MD, October 2003.
53
Hierarchy of Pharmacogenetic Information from
Single Nucleotide Polymorphisms (SNPs)
54
Hierarchy of Pharmacogenetic Information from
Single Nucleotide Polymorphisms (SNPs)
55
A whole genome approach can work
56
Current Methods for genetic testing

• By phenotype metabolic probe drug or Western
  blot
• By PCR with mutation-specific endonuclease
• By PCR and allele-specific hybrization
• By oligonucleotide chip hybridization
• By laser lithography - guided oligonucleotide
  chip hybridization.
• By rapid throughput pyrosequencing
• Taqman probe screening
• Whole genome chips
• Custom chips

Koch WH. Technology platforms for pharmacogenomic
diagnostic assays. Nat Rev Drug Discov. 2004
Sep3(9)749-61.
57
Homepage of www.pharmgkb.org
58
Browsing PharmGKB
59
Example of the PharmGKB gene variant browser
60
Estimated cost to the patient of Genetic Tests in
Clinical Practice

• By simple PCR for one mutation 10
• For 50 mutations 150
• By Chip for 20 mutations 70
• By Chip for 100 mutations 250

61
Ethical, legal and policy issues within
pharmacogenetics

• Risk of Loss of Patient Confidentiality
• Need for anonymized DNA storage systems
• Risk that existing patents will stifle progress
• Need for careful interpration of Bayh-Dole
• Untangling the relationship between genetics and
  self-described race and ethnicity

62
Role Models for Pharmacogenetics

• Concorde?
• Nuclear Power?
• The Longitude Problem?

63
Clinical PharmacogeneticsSummary

• A good phenotyping probe is critical
• Genetic tests need validation just as any other
  tests
• A potent inhibitor can mimick a genetic
  polymorphism
• Not all genetic polymorphisms have a phenotypic
  correlate, or clinical effect
• The clinical relevance of genetic polymorphisms
  is greatest with drugs of narrow therapeutic
  range, but not confined to them
• The cost of genetic testing is not likely to be
  limiting

64
Medication HistoryAVOID Mistakes

• Allergies? Is there any medicine that we should
  not give you for any reason?
• Vitamins and Herbs?
• Old drugs? ..as well as current
• Interactions?
• Dependence?
• Mendel Family Hx of benefits or problems with
  any drugs?

65
Pharmacogenetics Websites

• www.pharmgkb.org
• The SNP consortium http//brie2.cshl.org
• The Human Genome
• www.ncbi.nlm.nih.gov/genome/guide/H_sapiens.html
• CYP alleles www.imm.ki.se/CYPalleles/
• Drug Interactions www.drug-interactions.com