Clinical Pharmacogenetics David A Flockhart MD, PhD Chief, Division of Clinical Pharmacology Professor of Medicine, Genetics and Pharmacology Indiana.

Clinical Pharmacogenetics

David A Flockhart MD, PhD Chief, Division of Clinical Pharmacology

Professor of Medicine, Genetics and Pharmacology

Indiana University School of Medicine

Pharmacogenetics, 1990

Pharmacogenomic Journals, 2007

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 ethnicity

Role Models for Pharmacogenetics

• Concorde?

• Nuclear Power?

• The Longitude Problem?

SNP Variability in The Human Genome December 2007

• 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 > 1% frequency• ~450,000 SNPs in MCS (Multiply Conserved Regions)

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

SNPs that change clinical outcome

SNPs that change drug response

SNPs that change pharmacokinetics

SNPs that change activity in vitro

Non-conservative amino acid changes

Non-synonymous SNPs in exons

Exon-based changes

All SNPs

Hierarchy of Pharmacogenetic Information from Single Nucleotide Polymorphisms (SNPs)

Current Clinical Ability to Predict Response

Clinical

Value

of a

Pharmacogenetic

Test

Pharmacogenetic Principle 1:

Value Decreases when Current Predictive Ability is High

Meyer UA and Flockhart DA, 2005

Azathioprine/TPMT

β-blockade/β Receptor

Cancer Chemotherapy

Antidepressants/5HTR

Methods in Pharmacogenetics• SNP discovery:

– Candidate gene approach

– Pathway approach

– Genome Wide Arrays

• Identification of gene and variants• Development of a genetic test for DNA variants • Correlation between genotype and phenotype• Validation

• Application in Clinical Practice

Polymorphic Distribution

0

10

20

30

40

50

60

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Activity (Phenotype)

Fre

qu

en

cy

Antimode

The Value of Normit Distribution Plots:

Population Distribution of CYP2C19 phenotype

Flockhart et al: Clin Pharmacol Ther 1995;57:662-669

Skewed Distribution

0

10

20

30

40

50

60

1 2 3 4 5 6 7 8 9 10 11 12 13

Activity (Phenotype)

Fre

qu

ency

Example 1 of a Skewed Distribution: Heterogeneity in response to Inhaled

Corticosteroids

Weiss ST et al. Hum Molec Genetics 2004; 13:1353-1359

Using the extremes of a phenotypic distribution as a strategy to identify

pharmacogenetic predictors

Example: Iressa™ and the EGF receptor

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

Genetics and Drug Absorption

Enterocyte GI Lumen

ATP

ADP

P-gp Transport

Passive Diffusion

Digoxin Transport across the GI lumen

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).

P-Glycoprotein Pharmacogenetics :

Effect of a “wobble” (no coding change) SNP in exon 26

Eichelbaum et al. Proc Nat Acad Sci March, 2000.

Eichelbaum et al, Proc Nat Acad Sci, 2000:March

0.25 mg of digoxin po at steady state

Brain Blood

ATP

ADP

P-gp Transport

Passive Diffusion

Digoxin Transport across the Blood-Brain Barrier

Genetics and Drug Elimination

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

CYP2D6 Pharmacogenetics

120

80

40

00.01 0.1 1 10 100

1011 Subjects

Debrisoquine/4-HydroxydebrisoquineMetabolic Ratio

UMsEMs

PMs

cutoff

Nu

mb

er o

f S

ub

ject

s

New CYP2D6 variants continue to appear….

From: Zanger et al: Clin Pharmacol Ther. 2004 Aug;76(2):128-38.

CYP2D6 Alleles

• 67 as of December, 2007• 24 alleles have no activity• 6 have decreased activity• *1, *2, *4 and many others have

copy number polymorphisms• The *2 variant can have 1,2,3,4,5 or

13 copies i.e increased activity

Oligonucleotide array for cytochrome P450 genotesting

From: Flockhart DA and Webb DJ. Lancet End of Year Review for Clinical Pharmacology, 1998.

From: Dalen P, et al. Clin Pharmacol Ther 63:444-452, 1998.

Paroxetine and CYP2D6 genotype change the plasma concentrations of

endoxifen

Stearns, Flockhart et al. J Natl Cancer Inst. 2003;95(23):1758-64.

Endoxifen

0

10

20

30

40

50

60

70

Before After

En

do

xife

n (

nM

)

p = 0.004

4-OH Tamoxifen

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

Before After

4-O

H T

am

ox

ife

n (

nM

)

Inhibition of CYP2D6 Affects Endoxifen Concentrations

0

20

40

60

80

100

120

140

Jin Y et al: J Natl Cancer Inst 97:30, 2005

Wt/Wt, noinhibitor

Venlafaxine Sertraline Paroxetine *4/*4, noinhibitor

Plasma Endoxifen

(nM)

0

20

40

60

80

100

0 2 4 6 8 10 12

Relapse-free Survival

%%

Years after randomizationYears after randomization

P=0.009 (vs. 0.02 genotype only)P=0.009 (vs. 0.02 genotype only)

EMEM

IMIM

PMPM

2 year RFS: 2 year RFS:

EM 98%EM 98%

IM 92%IM 92%

PM: 68%PM: 68%

10 year NNT = 3

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

Vitamin K Carboxylase and CYP2C9 Genotypes altered Warfarin Dose

Rieder et al. N. Eng J. Med 2005;352: 2285-2293

Genetic alterations in Phase 2 enzymes with clinical consequences

UGT1A1 NAT-2

SULT1A1COMTTPMT

UGT1A1 TA repeat genotype altersirinotecan neutropenia/activity

35.7

16.3

8.6

0

5

10

15

20

25

30

35

40

45

50

6/6 6/7 7/7

P=0.007

UGT1A1 genotype

% g

rade

4/5

neu

trop

enia

N=524

41.9

33.8

14.3

05

1015202530354045

6/6 6/7 7/7

UGT1A1 genotypeO

bjec

tive

resp

onse

(%

)

P=0.045

McLeod H. et al, 2003.

N-Acetylation PolymorphismNAT-2

• Late 1940’s : Peripheral Neuropathy noted in patients treated for tuberculosis.

• 1959 : Genetic factors influencing isoniazid blood levels in humans. Trans Conf Chemother Tuberc 1959: 8, 52–56.

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

• Caffeine

• Dapsone

• Hydralazine

• Isoniazid

• Procainamide

Incidence of the Slow Acetylator NAT-2 phenotype

• 50% among Caucasians

• 50% among Africans

• 20% among Egyptians

• 15% among Chinese

• 10% among Japanese

0

20

40

60

80

100

120

0 20 40 60 80 100

Duration of Therapy (months)

% o

f p

ts w

ith

lu

pu

s

Slow Acetylators

Fast Acetylators

Onset of Positive ANA Syndrome with Procainamide.

Woosley RL, et al. N Engl J Med 298:1157-1159, 1978.

Thiopurine Methyl Transferase

• Homozygous mutants are 0.2% of Caucasian Populations

• Heterozygotes are ~ 10%

• Homozygous wild type is 90%– Metabolism of Azathioprine– 6-Mercaptopurine

Thiopurine Methyl Transferase Deficiency

From: Weinshilboum et al. JPET;222:174-81. 1982

Effect of TPMT genotype on duration of Azathioprine therapy.

From: Macleod et al: Ann Int Med 1998;

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

Ser Arg

Ser Gly

Gly Arg

Gly Gly

Ser Arg

Ser Arg

Ser Gly

Ser Arg

Gly Arg

Ser Arg

Gly Gly

Ser Arg

Gly Gly

Ser Gly

Gly Arg

Ser Gly

Ser Gly

Ser Gly

Gly Gly

Gly Arg

Gly Arg

Gly Arg

Haplotypes Diplotypes

Ying-Hong Wang PhD,

Indiana University School of Medicine

Gly Gly

Gly Gly

2SNPs: 10 possible hapoltypes

Observed 1AR Haplotypes in Caucasians and African American Women (WISE study)

Terra et al. Clin. Pharmacol. Ther. 71:70 (2002)

Ser Arg

Ser Arg

Ser Gly

Ser Arg

Gly Arg

Ser Arg

Gly Arg

Ser Gly

Gly Gly

Ser Arg

Gly Gly

Ser Gly

Ser Gly

Ser Gly

Gly Gly

Gly Arg

Gly Arg

Gly Arg

Haplotypes Diplotypes

SR/SR SR/SG SR/GR

SG/GR

Ser Arg

Ser Gly

Gly Arg

Gly Gly

Ying-Hong Wang PhD, Indiana University School of Medicine

Of 10 theoretical diplotypes, only 4 were present in the study population

Johnson et al. Clin Pharmacol & Ther. 2003,74:44-52.

Diplotype predicts Beta-blocker effect

Lesson: Diplotype may be a better predictor of effect than

Genotype

A SNP that tags a Haplotype (tagSNP) may be an economical

means of screening

Lanfear DE et al. JAMA September 28th, 2005;294:1526-1533.

β2 receptor Gln27Glu (79CG) genotype predicts mortality

during β-blockade after MI.

Non-synonymous coding region polymorphisms in long QT disease genes

NH2

NH2

NH2

COOH

extracellular

extracellular

intracellular

intracellular

COOH

COOH

NH2

NH2

COOH

COOH

SCN5AINa

IKs

IKr

KvLQT1

minK

HERG MiRP1

K897T(10%)

T8A(1.5%)

D85N(1.5%)

S38G(38%)

H558R(24%)

Vanderbilt

R34C (4%)

P1090L(4%)

G643S(9%)

Japan only

P448R(20%[?])

Dan Roden MD, October 2003.

A Genetic Effect on Hydrochlorothiazide Efficacy

SNPs that change clinical outcome

SNPs that change drug response

SNPs that change pharmacokinetics

SNPs that change activity in vitro

Non-conservative amino acid changes

Non-synonymous SNPs in exons

Exon-based changes

All SNPs

Hierarchy of Pharmacogenetic Information from Single Nucleotide Polymorphisms (SNPs)

SNPs that change clinical outcome

SNPs that change drug response

SNPs that change pharmacokinetics

SNPs that change activity in vitro

Non-conservative amino acid changes

Non-synonymous SNPs in exons

Exon-based changes

All SNPs

Hierarchy of Pharmacogenetic Information from Single Nucleotide Polymorphisms (SNPs)

Current Methods for genetic testing

• By phenotype: metabolic probe drug or Western blot or Immunohistochemistry

• 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

• By rapid and high throughput full sequencing

Medication History:AVOID Mistakes

Allergies? : Is there any medicine that we should not give you for any reason?

Vitamins and Herbs?Old drugs? …..as well as currentInteractions?Dependence?Mendel: Family Hx of benefits or problems

with any drugs?

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