Transcript Essai titre couleur

Genetic Epidemiology of
Breast Cancer
Susan Neuhausen, PhD
Epidemiology Div., Dept. Medicine
University of California Irvine
Irvine, CA USA
Risk factors for disease
Genotypes
Inherited (exposure)
Somatic (outcome)
Exposures
Physical and biologic agents, e.g., chemical exposures, drugs,
infectious agents
Lifestyle, e.g., smoking, alcohol use, nutrition, # pregnancies
Demographics
Interactions
Sex
Age
Ethnicity
Genotypes
Exposures
Demographics, e.g., age, sex
SEER US incidence rates for breast
cancer, 1988-1992
120
Non-Hispanic Cauc
Hispanic
African American
Alaska Native
Amer Indian
Chinese
Japanese
Vietnamese
Hawaiian
100
80
60
40
20
0
Rates per 100,000
Age-specific incidence of breast cancer for different population groups
Based on Cancer Incidence in
Five Continents
Familial risks
Family members
Increased risk
Personal history
1 first-degree relative
2 first-degree relatives
First-degree with bilateral
3-4 fold increased risk
2-fold
5-fold
6-fold
Importance of identifying genes
• Objectives
• To understand the etiology of the disease
• To develop more accurate models for risk
estimation
• Purpose
• To develop more specific strategies and
therapies
• To target individuals at highest risk of
developing cancer (or other diseases being
studied) for preventative strategies,
surveillance,etc.
General classes of genes causing
susceptibility to disease
Class
Variant
frequency
Absolute
risk
Attributable Examples
risk
Rare variants
Low
High
Low
BRCA1, BRCA2
Common
variants
High
Low
High
AR, NAT2
BRCA1 and BRCA2
3
9
10
12 5 6 7 8
BRCA1
11
Identified mutation
15
19 20 21
18
22
17
23
16
24
5382insC
BRCA2
10
14
13
185delAG
1 2 3 45 6 7 8 9
12
11
19
14 15 16 17 18
6174delT
20 21 22
23 25 26
24
27
Founder effect
A high frequency of a mutant gene in a population
founded by a small ancestral group
Ancestral group
One random mutation
Generations later
Mutation spread
throughout population
Examples of recurrent mutations
found in BRCA1 and BRCA2
Population
Mutation
Ashkenazi Jews
BRCA1 185delAG, 5382insC; BRCA2 6174delT
Icelanders
BRCA2 995delC
Dutch
BRCA1 IVS12-1643del3835; BRCA2 5573insA
Norwegians
BRCA1 1136insA and 1675delA
Swedes
BRCA1 3171ins5,
Poles
BRCA1 5382insC; C61G, 4153delA
What factors modify the risk of
developing cancer, the type of cancer, and
the age at which it develops?
•
Genotype-phenotype correlations
• OCCR region in BRCA2 and BRCA1
•
Gene-environment interactions
• Oral contraceptive use
• Parity, Age at menarche
• Prophylactic oophorectomy
•
Gene-gene interactions
• Androgen receptor
• AIB1
• RAD51
Breast and ovarian cancer risks by location
From Thompson & Easton, J Mammary Gland Biol Neoplasia 9:221. 2004
Age specific penetrance differs by ascertainment
From Thompson &
Easton, J Mammary
Gland Biol Neoplasia
9:221. 2004
There is substantial inter-individual
variability in cancer risk among BRCA1
and BRCA2 mutation carriers
• Even among those carrying the same
deleterious mutation, there is large variation in
age at diagnosis and cancer type
• This suggests that there are additional genetic
and environmental factors that modify risk
Collaborative group of BRCA1/2 mutation
carriers for modifier studies
Baylor-Dallas
Baylor-Houston
City of Hope
Creighton
Dana Farber
Duke
Fox Chase
Georgetown
Mayo
Northwestern
UChicago
UC Irvine
UCLA
UPenn
UT Southwestern
UToronto
UVienna
Candidate Modifiers:
IGF (UC Irvine)
BRCA1/2
Cohort
2434 carriers and
1600DNAs**
Candidate Modifiers:
DNA repair (UPenn)
Candidate Modifiers:
TBA
Novel Modifiers
Detection (Penn)
Breast cancer families due to
BRCA1 and BRCA2
Family group
BRCA1
BRCA2 Other
All Families 4+ BrCa (n=237)
52%
32%
16%
4-5 BrCa (female only) (n=83)
28%
5%
67%
6+ BrCa (female only) (n=34)
21%
60%
19%
Br and Ovarian Ca (n=94)
81%
14%
5%
Male BrCa in family (n=26)
16%
76%
8%
From Ford et al., 1998
Efforts to find additional high penetrance
breast cancer susceptibility genes
• Study families known not to have BRCA1 or
BRCA2 mutations
• No ovarian or male breast cancer in the family
• At least three cases of breast cancer diagnosed <60
years
• Helpful to stratify by
•
•
•
•
Gene expression
Comparative genomic hybridization
Tumor histology
Other cancer types
Putative BRCA3 loci
• 8p12-p22
• Seitz et al., 1997; 3-point LOD = 3.30 in 2 German families
• Rahman et al., 2000; 31 BC families. No linkage evidence
• 13q21
• Kainu et al., 2000. Scandanvian families. 2-point LOD of 2.8;
multipoint HLOD of 3.46. Tumors had LOH at 13q21-q22.
• Thompson et al., 2002. 128 families- no linkage evidence
• Other linkage evidence from BCLC: 149 families w/>3
BC cases
• Best LOD of 2.4 on 2p in families with 4+ cases <age 50
• CONCLUSION: No gene is likely to account for a large
fraction of the breast cancer families. Likely multiple
genes with small contributions
Polygenic model of cancer risk
0.45
0.4
0.35
DENSITY
0.3
Cases
0.25
0.2
Population
0.15
0.1
0.05
0
0.001
0.01
0.1
1
RISK
10
100
Association studies
• A marker locus is associated with a disease if
the distribution of genotypes at the marker
locus in disease-affected individuals differs
from the distribution in the general
population
• A specific allele may be positively associated
(over-represented in affecteds) or negatively
associated (under-represented)
CHEK2 1100delC and breast cancer
•
•
•
•
CHEK2 encodes a G2/M checkpoint kinase.
1100delC variant abolishes the kinase activity
1.4% frequency in controls in Europe; 0.4% in US
Meijers-Heijboer and BrCa Consortium first reported an
association with breast cancer
• 2-fold increased risk for breast cancer
• Attributable risk of 9% for male breast cancer, based on families
with male breast cancer
• Vajterostp et al. 2002 confirmed 2-fold increased risk BC
• Male breast cancer: Neuhausen et al., 2004; Syrjakoski et
al., 2004 found no increased risk
• CONCLUSION: 2-fold increased risk in women with a family
history
ATM and breast cancer risk
• Carriers of mutations in ATM suffer from
ataxia-telangiectasia (AT)
• Protein is activated in response to DNA
damage by ionizing radiation
• Female heterozygotes are at a 3-4 fold
increased risk of breast cancer
• Inconclusive results of ATM missense
mutations and breast cancer risk
Summary of six breast cancer
susceptibility genes
Gene
Pop’n carrier
frequency
Risk of BC by 70
yrs (95% CI)
BRCA1
BRCA2
CHEK2
ATM
1 in 860
1 in 740
1 in 90
1 in 100
65% (44-78%)
45% (31-56%)
11% (9-14%)
23% (13-39%)
TP53
PTEN
1 in 5,000
1 in 250,000
50-60% by 45 yrs
30-50%
Other associations have been reported but not replicated
From Thompson & Easton. 2004. J Mammary Gland Biol Neoplasia 9:221.
Contribution of known genes to
breast cancer
BRCA1
BRCA2
TP53,PTEN, STK11
ATM/CHEK2
Other genes,
familial not genetic
Association studies to identify additional low
to moderate penetrance genes
• Need large numbers
• Genome-wide using 100+K SNPs
• Targeted pathways, e.g.
•
•
•
•
Insulin-like growth factor
Estrogen metabolism
Inflammation
DNA repair
• Incorporation of additional factors into models
for analysis
• Lifestyle factors
• Gene x gene interactions
• Gene x environment interactions
Considerations for association studies
• Study Design: case-control, cohort, familybased
• Choice of cases
• Choice of controls
• Choice of genes and pathways
• Selection of genetic variants
• Statistical issues
Multi-Step Carcinogenesis
A 21th Century View: