Environment and Breast Cancer: Science Review

Evidence From Humans
Print this page
Polymorphisms in DNA repair genes, traffic-related polycyclic aromatic hydrocarbon exposure and breast cancer incidence
Mordukhovich, I., Beyea, J., Herring, A. H., Hatch, M., Stellman, S. D., Teitelbaum, S. L., Richardson, D. B., Millikan, R. C., Engel, L. S., Shantakumar, S., Steck, S. E., Neugut, A. I., Rossner, P., Jr., Santella, R. M., Gammon, M. D. Int J Cancer. 2016. 139:2, 310-21.
Topic area
Environmental pollutant - PAHs Air pollution, vehicle exhaust
Study design
Population-based case-control
Funding agency
U.S. Department of Defense NCI NIEHS Breast Cancer
Study Participants
Number of Cases
For 1995 exposure analyses: 842-905 cases For 1960-1990 exposure analyses: 332-429 cases
Menopausal Status
The menopausal status of women included in this study is listed here.
No analysis based on menopausal status
Number of Controls
For 1995 exposure analyses: 911-958 controls For 1960-1990 exposure analyses: 368-474 controls
Participant selection: Inclusion and exclusion criteria
Criteria used to select participants in the study.
Female residents of Nassau and Suffolk Counties (Long Island), NY, participating in the Long Island Breast Cancer Study Project, age 20 or older, English-speaking, newly diagnosed with in situ or invasive breast cancer in 1996-1997. Cases identified by regional hospital pathology laboratories. Controls had no breast cancer history and were matched by 5-year age group, identified by random-digit-dialing or Medicare records (for women 65 and older). Approximately 1102 cases (73.1% of LIBCSP cases) and 1140 controls (73.3% of LIBCSP controls) provided a blood sample; the analyses reported here were limited to women for whom samples could be sufficiently genotyped. The main reason that samples could not be genotyped (<10%) was insufficient DNA. Also, when residential histories were incomplete, imputation methods were used, and analyses of the 1960-1990 exposure data were limited to participants with ≤ 20% imputed exposure data. These restrictions are reflected in the number of cases/controls reported above. The number of cases and controls per analysis also varied depending on the particular polymorphism being analyzed.
Comment about participation selection
In the LIBCSP, giving a blood sample was positively associated with being white, ever using alcohol, ever using HRT, ever having a mammography, and lactation history. Older women and former smokers were less likely to give blood. Blood donation was not associated with case-control status, so these differences between the total study population and the sub-population who donated blood should not bias the findings, but could affect generalizability.
Exposure Investigated
Exposures investigated
Geocoded self-reported residential histories in Nassau and Suffolk counties were used to model recent (1995) and long-term (1960-1990) ambient B[a]P from traffic emissions. Emissions were estimated using hourly roadway-specific traffic counts and average
How exposure was measured
Questionnaire, in-person Biological GIS/geographic location
Exposure assessment comment
Tailpipe emissions for years other than 1960, 1970, 1980 and 1990 were interpolated or extrapolated. The exposure model was validated against measurements of PAHs in soil at participant residences as well as levels of PAH-DNA adducts in blood. Limitations include not accounting for a street canyon effect, historical changes to the road network, industrial emissions, and exposure that occurred away from the home. In analysis of exposure from 1960-1990, imputation methods were used to estimate exposure for years when participants were missing residential history or for years lived outside of study area. Analyses were restricted to individuals with ≤ 20% of their cumulative exposure based on imputed values. Short-term (1995) and long-term (1960-1990) exposure estimates were strongly correlated. Most participants had lived in current homes for at least 15 years. Genes in this analysis are all involved in nucleotide excision repair or base excision repair. ERCC2 is also known as XPD. Previous papers from the LIBCSP (e.g. Terry et al, 2004) referred to this gene as XPD, so we do so in reporting strengths of association below.
Breast cancer outcome investigated
Primary incident breast cancer
Confounders considered
Other breast cancer risk factors, such as family history, age at first birth, and hormone replacement therapy use, that were taken into account in the study.
Based on causal diagram: Education, annual household income, race, religion, parity, age at first birth, BMI, duration of oral contraceptive use, lifetime average alcohol intake, lifetime average physical activity, and duration of breastfeeding
Genetic characterization included
If the study analyzed relationships between environmental factors and inherited genetic variations, this field will be marked Yes. No, if not.
Strength of associations reported
Tertile 3 vs Tertile 1 for recent residential traffic-related PAH exposure estimates (1995):

XPD* Lys751Gln, AA wild-type: aOR 0.86 (95% CI 0.60-1.24)
XPD Lys751Gln, AC variant: aOR 0.94 (95% CI 0.68-1.31)
XPD Lys751Gln, CC variant: aOR 2.09 (95% CI 1.13-3.90)
p-interaction = 0.04

XRCC1 Arg399Gln, GG wild-type: aOR 0.97 (95% CI 0.68-1.37)
XRCC1 Arg399Gln, GA variant: aOR 1.07 (95% CI 0.77-1.48)
XRCC1 Arg399Gln, AA variant: aOR 0.95 (95% CI 0.48-1.88)
p-interaction = 0.90

XRCC1 Arg194Trp, CC wild-type: aOR 0.96 (95% CI 0.76-1.22)
XRCC1 Arg194Trp, CT or TT variant: aOR 1.44 (95% CI 0.76-2.75)
p-interaction = 0.54

OGG1 Ser326Cys, CC wild-type: aOR 1.05 (95% CI 0.78-1.40)
OGG1 Ser326Cys, GC variant: aOR 0.94 (95% CI 0.64-1.39)
OGG1 Ser326Cys, GG variant: aOR 1.11 (95% CI 0.38-3.23)
p-interaction = 0.78

0-1 high risk alleles: aOR 0.88 (95% CI 0.49-1.57)
2-3 high-risk alleles: aOR 0.92 (95% CI 0.70-1.20)
4-6 high-risk alleles: aOR 2.32 (95% CI 1.22-4.49)

Tertile 3 vs Tertile 1 for long-term residential traffic-related PAH exposure estimates (1960-1990):

XPD*Lys751Gln, AA wild-type: aOR 1.19 (95% CI 0.66-2.12)
XPD Lys751Gln, AC variant: aOR 1.27 (95% CI 0.74-2.16)
XPD Lys751Gln, CC variant: aOR 2.24 (95% CI 0.80-6.25)
p-interaction = 0.50

XRCC1 Arg399Gln, GG wild-type: aOR 1.88 (95% CI 1.04-3.41)
XRCC1 Arg399Gln, GA variant: aOR 1.02 (95% CI 0.61-1.72)
XRCC1 Arg399Gln, AA variant: aOR 1.07 (95% CI 0.36-3.18)
p-interaction = 0.59

XRCC1 Arg194Trp, CC wild-type: aOR 1.17 (95% CI 0.79-1.73)
XRCC1 Arg194Trp, CT or TT variant: aOR 3.04 (95% CI 0.97-9.51)
p-interaction = 0.30

OGG1 Ser326Cys, CC wild-type: aOR 1.77 (95% CI 1.09-2.88)
OGG1 Ser326Cys, GC or GG variant: aOR 0.84 (95% CI 0.48-1.47)
p-interaction = 0.52

0-1 high risk alleles: aOR 0.89 (95% CI 0.36-2.22)
2-3 high-risk alleles: aOR 1.25 (95% CI 0.80-1.95)
4-6 high-risk alleles: aOR 2.96 (95% CI 1.06-8.21)

*XPD is another name for ERCC2
Results Comments
There was also evidence for additive interaction for XPD Lys751 Gln and 1995 traffic-related PAH exposure; the relative excess risk due to interaction (RERI) was 0.32 (0.074-0.57). No evidence of interaction between PAH exposure from traffic and the following DNA repair polymorphisms: XPA-4A/G, ERCC4 Arg415Gln, ERCC5 Asp1104His, XPD (also called ERCC2) Asp312Asn and ERCC1 8092C/A. After considering linkage disequilibrium between polymorphisms, the sum of "high-risk alleles" included: XPD Lys751Gln, OGG1 Ser326Cys and XRCC1 Arg194Trp.
Author address
Department of Epidemiology, Lambertville, NJ. Consulting in the Public Interest, Lambertville, NJ. Biostatistics, University of North Carolina, Chapel Hill, NC. Carolina Population Center, University of North Carolina, Chapel Hill, NC. Division of Cancer
Controls participation rate
63% completed interview 46% both completed intervi