Genetic Variation in Seven Obesity Genes and Risk of Postmenopausal Breast Cancer

1
Genetic Variation in Seven Obesity Genes and Risk
of Postmenopausal Breast Cancer

• Heather Spencer Feigelson
• Kaiser Permanente
• Denver, Colorado

http//www.gcarlson.com
2
Obesity and Breast Cancer

• Obesity is a leading risk factor for
  postmenopausal breast cancer
• This increased risk results primarily from
  conversion of androstenedione to estrone in
  adipose tissue
• Adipose tissue is an active endocrine organ, and
  thus may also play a role in tumor development
  independent of estrogen production

3
Hypothesis

• Variation in genes that encode adiposity-related
• proteins my influence risk of breast cancer
• Candidate genes were chosen that had been
  associated with obesity
• and
• 1) had previously been associated with breast
  cancer or another cancer
• 2) or had demonstrated activity in breast cancer
  cell lines

4
Candidate Genes
5
Methods

• Study Design
• Nested case-control study
• Study Population
• American Cancer Society CPS-II Cohort
• 648 postmenopausal breast cancer cases
  (diagnosed 1992-2001)
• 659 cancer-free controls
• matched on age, ethnicity, date of blood
  collection
• 99 white
• Mean age at blood draw 69

6
SNP Selection

• Database of genetic variants (www.HapMap.org)
• Selected tagging SNPs that capture common
  variation
• Minor allele frequency 5

7
SNP Selection
A total of 72 SNPs in 7 genes were identified
Use linkage disequilibrium (LD) to reduce SNP
number (pair-wise correlation of r2gt0.80)
Choose 45 tagging SNPs
Assays designed for SNPstream genotyping system
(Emory University)
39 SNPs genotyped
1 SNP dropped for HWE
8
Statistical Analysis

• Unconditional logistic regression
• Controlled for matching factors
• Also examined possible confounders and tested for
  effect modification
• BMI, weight gain, PMH use, family history,
  history of breast cysts
• Haplotype analysis for statistically significant
  SNPs in LD

9
Log additive model p-values for the association
between breast cancer and 38 SNPs
10
Pattern of LD for tagging SNPs genotyped in
HSD11B1
11
HSD11B1 SNP and Haplotype Associations with
Postmenopausal Breast Cancer
12
Pattern of LD for tagging SNPs genotyped in IRS2
13
IRS2 SNP and Haplotype Associations with
Postmenopausal Breast Cancer
14
Additional Results

• No other SNPs were associated with breast cancer
• No evidence of effect modification for any SNP
• BMI, PMH, family history, history of breast cyst

15
Conclusions

• Common variation in HSD11B1 and IRS2 may be
  associated with breast cancer among
  postmenopausal white women
• Previous studies of IRS2 have been inconsistent
• Some have shown associations with obesity and
  others with breast cancer
• One SNP in a recent GWAS showed modest
  association with breast cancer (0.051), but this
  SNP was not associated with breast cancer in our
  data

16
Conclusions

• HSD11B1 has not been previously studied as a
  candidate gene for breast cancer, however,
  genome-wide studies of breast cancer have not
  seen an association
• HSD11B1 has been previously associated with type
  2 diabetes, hypertension, PCOS
• HSD11B1 possible candidate gene for weight
  change (GWAS)

17
Conclusions

• Strengths of this study
• Population-based design
• Use of tagging SNPs to comprehensively capture
  variation
• Limitations of this study
• Small sample size
• Included only postmenopausal women
• 99 white
• Confirmation of these findings in a larger study
  is needed

18
Acknowledgements

• American Cancer Society (Atlanta, GA)
• Lauren Teras, Ryan Diver, Alpa Patel, Jeanne
  Calle, Vicky Stevens, and Michael Thun
• Emory University Center for Medical Genomics
• Mark Bouzyk and Weining Tang
• Citation Breast-Cancer-Research.com/content/10/4/
  R57

19
(No Transcript)
20
Biosynthesis of Estrogens in Postmenopausal Women
21
Modifiable Risk Factors for Postmenopausal
Breast Cancer
RR2.0
RR1.6
RR1.4
RR1.2
RR1.2
RR1.1
No association
Relative Risk
22
The American Cancer Society Cancer Prevention
Study-II
Repeat questionnaires Nutrition Cohort
1997 1999 2001 03 05
1992 Baseline Questionnaire
1982 Baseline Questionnaire
2002
1998
Nutrition Cohort
Mortality Cohort
Blood
1.2 million
184,000
40,000
57 Women 97 White Age 30-108 (median52)
Buccal Cell
70,000