No Association between Ovarian Cancer Susceptibility Variants and Breast Cancer Risk among Chinese Women

Breast and ovarian cancer are hormonally driven and have similar etiologies, including inherited mutations in BRCA1 or BRCA2 (1). We hypothesized that these cancers may share common genetic susceptibility variants. To date, genome-wide association studies (GWAS) have identified 10 ovarian cancer susceptibility variants; these studies have included only Caucasian women (2). To the best of our knowledge, no previous studies have evaluated the association between ovarian cancer susceptibility variants and breast cancer risk.

This analysis included a total of 2,918 breast cancer cases from the Shanghai Breast Cancer Study (SBCS), Shanghai Breast Cancer Survival Study (SBCSS), and the Shanghai Women's Health Study (SWHS), and 2,324 controls from the SBCS and SWHS. Detailed study design and data collection procedures have been described previously (3, 4). Study protocols were approved by the relevant review boards of all institutions, and informed consent was obtained from all participants.

Genetic variants associated with ovarian cancer susceptibility were identified in the GWAS catalog (2). Linkage disequilibrium among Han Chinese was assessed using SNAP (5). Independence of loci was defined using an r² ≥ 0.6. Genotyping and imputation methods and quality control have been previously described (3). Briefly, genotyping data were from the Affymetrix Genome-Wide Array (6.0); only variants with quality control values of 0.95 or more were included. Imputation was conducted using MACH; only data with quality scores [r-squared correlation coefficient (RSQ)] >0.3 (mean 0.98, median 1.00) were included. Furthermore, only common genetic variants [minor allele frequency (MAF) ≥0.05] were evaluated.

Multivariate logistic regression was used to estimate ORs and 95% confidence intervals (CI) for associations between breast cancer risk and genetic variants using additive effect models that included adjustment for age and education. Effect measure modification was evaluated with stratified analyses and the likelihood ratio test. Quanto was used for power calculations (6). All other analyses were conducted with SAS version 9.2 (SAS institute Inc.). All statistical tests were two-tailed, and statistical significance was defined by a P value 0.05 or less.

As shown in Fig. 1, in addition to the 10 original ovarian cancer susceptibility variants in the GWAS catalog, 353 additional variants in linkage disequilibrium (r² ≥ 0.6) among Han Chinese were identified. Genotyped (N = 51) or imputed (N = 141) data were available for 192 variants; however, 89 were excluded from analysis (MAF < 0.05). Thus, 25 genotyped and 78 imputed common (MAF ≥ 0.05) variants, including 5 original and 98 additional ovarian cancer susceptibility variants were evaluated. No associations with breast cancer risk in additive effect models adjusted for age and education were identified (full results in Supplementary Table S1; summary findings in Table 1). These 103 variants represent 5 of 9 loci identified. Three loci were not evaluated as they included only uncommon variants (MAF < 0.05). One locus was not evaluated as it included only 1 variant (rs8170), for which neither genotyping nor imputed data were available. However, according to HapMap, this variant also has a very low MAF among Han Chinese (0.004).

Stratified analyses by menopausal status, estrogen receptor (ER), progesterone receptor (PR), and tumor stage were conducted to evaluate effect measure modification. Two variants in locus 7 (rs7207826 and rs136870) were associated with increased risk among ER+ tumors, whereas 1 variant in locus 5 (rs1416745) was associated with a decreased risk among late stage tumors. However, these nominally significant associations did not withstand correction for multiple comparisons.

In this large, population-based, case–control study, we did not observe any significant associations between ovarian cancer susceptibility variants and breast cancer risk. To the best of our knowledge, this is the first study aimed to evaluate the association between ovarian cancer susceptibility variants and breast cancer risk. Two studies have evaluated the association between breast cancer susceptibility variants and ovarian cancer risk (7, 8). The first included 7 variants and found no association (7). The second included 11 variants and found a significant association between rs4954956 and ovarian cancer risk (OR, 1.07; 95% CI, 1.01–1.13; ref. 8).

Strengths of this study include the large study population, the use of both genotyped and imputed data to maximize genetic coverage, and sufficient power to detect associations for common genetic variants. For a variant with a MAF of 20%, this study had greater than 78% power to detect an OR of 1.15. A limitation of this study is that rare variants (MAF < 5%) were not evaluated, as very large studies are needed for adequate power. Notably, of the 10 original ovarian cancer susceptibility variants, 4 had low MAFs among Han Chinese (rs2665390, rs10088218, rs12794435, and rs8170). Therefore, differences in the genetic architecture of ovarian cancer between Chinese and Caucasian populations may influence our results. Notably, all 10 ovarian cancer susceptibility variants identified to date are from studies of Caucasian women.

In conclusion, we found no evidence for an association between ovarian cancer susceptibility variants and breast cancer risk among Chinese women. Resequencing and fine-mapping of ovarian cancer susceptibility loci for evaluation of rare variants may be necessary to fully evaluate associations with breast cancer risk.

No potential conflicts of interest were disclosed. The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Conception and design: X. Ma, X.-O. Shu, W. Lu, Y.-T. Gao, W. Zheng, A. Beeghly-Fadiel

Development of methodology: W. Lu, W. Zheng

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Q. Cai, X.-O. Shu, Y.-T. Gao, W. Zheng

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): X. Ma, R.J. Delahanty, B. Zhang, J. Long, A. Beeghly-Fadiel

Writing, review, and/or revision of the manuscript: X. Ma, Q. Cai, X.-O. Shu, Y.-T. Gao, W. Zheng, A. Beeghly-Fadiel

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y.-T. Gao, W. Zheng

Study supervision: X.-O. Shu, W. Lu, Y.-T. Gao, W. Zheng

The authors thank the participants and research staff of the SBCS, SBCSS, and the SWHS for their contributions and commitment to this project, Regina Courtney for DNA preparation, and Samantha Stansel for assistance with the submission of this article.

This research was supported by grants from the NIH/National Cancer Institute (NCI), including R01CA124558, R01CA64277, and R01CA90899. A. Beeghly-Fadiel is supported in part by a grant from the NIH/NICHD (5K12 HD043483-09). DNA extraction and genotyping was conducted at the Survey and Biospecimen and Functional Genomic Resource Laboratory, at the Vanderbilt Microarray Shared Resource, which is supported in part by the Vanderbilt Ingram Center (P30CA68485). X. Ma (from the Third Military Medical University in China) is supported by the China Scholarship Council (CSC) while the author is visiting the Vanderbilt Epidemiology Center in the United States.