Abstract
Genome-wide association studies (GWAS) have identified several loci as being associated with breast cancer in mostly European populations. We focus on TNRC9 rs3803662, FGFR2 rs1219648 and rs2981582, MAP3K1 rs889312, and 2q35 rs13387042, to replicate in the 4-Corner’s Breast Cancer Study of Hispanic (N = 565 cases and 714 controls) and non-Hispanic white (NHW) women (N = 1177 cases and 1330 controls). We evaluate associations by ethnicity, menopausal status, and tumor ER/PR status after adjusting for genetic admixture. TNRC9 AA genotype was associated with significant increased risk among NHW women (OR 1.54, 95% CI 1.14, 2.08; P trend 0.003). Both polymorphisms of FGFR2 were associated with statistically significant increased risk for NHW and Hispanic women; MAP3K1 was not associated with risk among either ethnic group. The polymorphism on 2q35 was associated with a statistically significant increased risk among Hispanic women (OR 1.53, 95% CI 1.08, 2.15 for the AA genotype; P trend = 0.004). Associations were significantly different among pre/peri-menopausal women for TNRC9 (P heterogeneity 0.008) and for 2q35 (P heterogeneity 0.08) for NHW and Hispanic women. Both FGFR2 polymorphisms reduced risk of ER−/PR− tumors in the presence of the minor allele among NHW women. Among Hispanic women, polymorphisms of the FGFR2 gene were associated with almost a twofold increase risk of an ER+/PR+ tumor, while non-significantly inversely associated with ER−/PR− tumors. Our data replicated some of the previously reported GWAS findings. Differences in associations were detected for NHW and Hispanic women by menopausal status and by ER/PR status of tumors.
Keywords: Breast cancer, Hispanic, TNRC9, MAP3K1, FGFR2, 2q35
Introduction
Genome-wide association studies (GWAS) of breast cancer have primarily focused on women of European ancestry [1–3]. These studies have identified new loci for replication and further evaluation. Replication studies in African American populations have had mixed success in confirming associations [4]. Among women of Asian ancestry approximately half of loci identified in women of European ancestry could be replicated [5]. Lack of replication of some loci could stem from differences in populations studied that varied by menopausal status, family history status, ethnicity, and other unique population characteristics. Confirmatory findings have been shown for a polymorphism in the fibroblast growth factor receptor 2 (FGFR2) gene, rs1219648 and in 2q35 rs13387042 in African American women [6]. This FGFR2 polymorphism was originally reported by Hunter [2] while the 2q35 polymorphism was reported by Stacey et al. [7]. Two polymorphisms in the FGFR2 gene, rs1219648 and rs2420946, were originally identified by Hunter among post-menopausal breast cancer cases [2]. A GWAS conducted by Easton and colleagues also identified a significant association with an FGFR2 with breast cancer using marker rs2981582 [1]. The finding reported by Hunter was confirmed in three of the four populations assessed. In 2007, Easton also reported on a polymorphism in MAP3K1, rs889312, which has been confirmed in other populations and has been identified as involved in a potential key pathway for breast cancer [1, 3]. TNRC9, also known as Tox3, was originally identified as important by Easton and the rs3803662 polymorphism has recently been confirmed from deep sequencing studies as a key TNRC9 SNP associated with breast cancer [8].
In this paper, given the confirmatory results from several studies in multiple populations, we focus on five polymorphisms, TNRC9 rs3803662, FGFR2 rs1219648 and rs2981582, MAP3K1 rs889312, and 2q35 rs13387042, to replicate in a breast cancer case–control study among women living in the Southwestern United States. Replication of these findings from GWAS in Hispanic women has not been published. We use data collected from the 4-Corner’s Breast Cancer Study conducted in Arizona, Colorado, New Mexico, and Utah to evaluate these associations. We consider differences in association that may stem from population characteristics including having a family history of breast cancer, menopausal status, and ER/PR status of tumors. We adjust for genetic admixture among Hispanic women.
Methods
Study participants were women living in Cochise, Coconino, Maricopa, Pima, Pinal, Santa Cruz, and Yuma Counties in Arizona, or the states of Colorado, New Mexico, or Utah at the time of diagnosis or selection [9]. Study hypotheses focused specifically on breast cancer in Hispanic women, therefore, sampling was stratified on ethnicity to select these women in larger proportion than their representation in the population. All Hispanic women between the ages of 25 and 79 who were diagnosed with breast cancer during the study period were eligible for the study. An age-matched sample of non-Hispanic white (NHW) women were randomly selected on a 1 to 1 ratio to the distribution of Hispanic cases in Arizona and Colorado; at a 4 to 1 ratio to the distribution of Hispanic cases in Utah; all Hispanic and non-Hispanic cases age 50 and under in New Mexico; and a 1 to 1 ratio for women over 50 in New Mexico. The GUESS program (Generally Useful Ethnic Search System) was used to identify women who were Hispanic [10]. A detailed description of study methods and response rates has been published and are briefly described below [9].
Cases were histologically confirmed as having in situ or invasive breast cancer (ICDO sites C50.0–C50.6 and C50.8–C50.9) diagnosed between October 1999 and May 2004. State tumor registries were used to initially identify and confirm case eligibility. They also provided study data for ER and PR status of tumors. Of cases identified, 68% of women contacted participated. Of the cases participating, 565 Hispanic and 1177 non-Hispanic white cases had DNA available for analyses.
Controls were selected from the target populations to match ethnicity and 5-year age distribution of cases. In Arizona and Colorado, participants under 65 were randomly selected from a commercial mailing list; in New Mexico and Utah, controls under 65 years were randomly selected from drivers’ license lists. In all states, women 65 years and older were randomly selected from Center for Medicare Services lists. Of controls identified 714 Hispanic and 1330 non-Hispanic white controls had DNA available for analyses.
All women identified were screened for eligibility prior to study enrollment. As part of the screening, women were asked to self-identify their race and ethnicity. Women initially identified as being Hispanic by the GUESS program who subsequently self-reported not being Hispanic/AI were ineligible for the study. All participants signed informed written consent prior to participation; the study was approved by the Institutional Review Board for Human Subjects at each institution. Respondents were given the option of having the interview administered in either English or Spanish. At the time of interview, respondents were again asked to self-identify their ethnicity and race as part of the study questionnaire. If a respondent described herself as belonging to more than one race or ethnic group, all were recorded. The questionnaire included information family history of first-degree relatives with breast cancer. As part of the interview, women were asked to “best describe your menstrual status on (referent date)” by selecting response from a card; this information was used to define individual menopausal status.
Genotyping
Genomic DNAs were isolated from blood from all study subjects. TaqMan assays for TNRC9 rs3803662, FGFR2 rs1219648 and rs2981582, MAP3K1 rs889312, and 2q35 rs13387042 were purchased from Applied Biosystems (Foster City, CA) and run according to the manufacturer. Briefly, each 5 µl TaqMan reaction contained 20 ng of genomic DNA, primers, probes, and TaqMan Universal Master Mix (containing AmpErase UNG, AmpliTaq Gold polymerase, dNTPs, and reaction buffer). TaqMan reactions were carried out with the following procedure: 50°C for 2 min to activate UNG, 95°C for 10 min, followed by 40 cycles of 92°C for 15 s, and 60°C for 1 min in an ABI 9700 PCR machine with 384 well format. Fluorescent endpoints of the TaqMan reactions were measured using an ABI 7900HT sequence detection instrument. All markers were in Hardy Weinberg equilibrium for both ethnic groups.
Statistical methods
All analyses were performed using SAS version 9.2 (SAS Institute, Cary, NC). The five loci were identified from GWAS studies. They were tested for Hardy–Weinberg equilibrium (HWE) within controls by ethnicity, and both R-square and D prime were calculated for the two loci in FGFR2 within controls by ethnicity using the EM algorithm to estimate haplotype frequencies. A chi-square test was used to compare the minor allele frequencies between ethnicities within controls. Multiple logistic regression models were used to assess breast cancer risk associated with the five loci, and analyses were stratified by ethnicity and adjusted for age at diagnosis/selection, study center and genetic admixture among Hispanic women. Methods for calculating genetic admixture have been described and were based on a two population model that included European and Native ancestry using the program Structure [11]. P values of linear dose response genotype effect were calculated by comparing the difference in the maximum likelihood estimates for a logistic regression model with and without the genotype of interest as ordinal using a chi-square test with one degree of freedom. Heterogeneity P values for differences in association between NHW and Hispanic women were determined by comparing the difference in maximum likelihood estimates for a logistic regression model with and without an interaction term between race and the genotype of interest using a chi-square test with one degree of freedom. We presented findings stratified by pre/peri-menopausal women as determined from data reported on the study questionnaire. We present findings stratified by ER and PR status of tumors for NHW and Hispanic women separately. No differences in association were detected by family history of breast cancer in first-degree relatives therefore those data are not presented.
Results
Study population characteristics are shown in Table 1. Most study participants were between 40 and 59 years of age and were post-menopausal. A family history of breast cancer in first-degree relatives was reported by 15.8% of NHW controls and 13.6% of Hispanic controls. Hispanic cases had more ER negative tumors than NHW women, although PR tumor status did not differ by ethnicity. The minor allele frequency (MAF) for the TNRC9 and MAP3K1 variants was significantly greater among Hispanic women, while the minor allele of the polymorphism at 2q35 was more common among NHW women.
Table 1.
Description of study population of women living in the Southwestern United States
| White, non-Hispanic | Hispanic | χ2 P valuea |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Cases | Controls | Cases | Controls | ||||||
| N | % | N | % | N | % | N | % | ||
| Age | |||||||||
| 25–39 | 74 | 6.3 | 101 | 7.6 | 58 | 10.3 | 82 | 11.5 | |
| 40–49 | 348 | 29.6 | 347 | 26.1 | 193 | 34.2 | 198 | 27.7 | |
| 50–59 | 333 | 28.3 | 351 | 26.4 | 162 | 28.7 | 194 | 27.1 | |
| 60–69 | 283 | 24.1 | 300 | 22.6 | 106 | 18.8 | 163 | 22.8 | |
| 70–79 | 137 | 11.7 | 230 | 17.3 | 46 | 8.1 | 79 | 11.0 | |
| Center | |||||||||
| AZ | 165 | 14.0 | 263 | 19.8 | 116 | 20.5 | 161 | 22.5 | |
| CO | 237 | 20.2 | 218 | 16.4 | 108 | 19.1 | 132 | 18.4 | |
| NM | 464 | 39.5 | 495 | 37.2 | 248 | 43.9 | 245 | 34.2 | |
| UT | 309 | 26.3 | 353 | 26.6 | 93 | 16.5 | 178 | 24.9 | |
| Menopause status | |||||||||
| Pre/Peri | 414 | 35.3 | 413 | 31.1 | 230 | 40.9 | 260 | 36.4 | |
| Post | 758 | 64.7 | 915 | 68.9 | 333 | 59.1 | 454 | 63.6 | |
| Family history of breast cancer | |||||||||
| No | 879 | 77.0 | 1068 | 84.2 | 429 | 81.3 | 568 | 86.5 | 0.05 |
| Yes | 263 | 23.0 | 201 | 15.8 | 99 | 18.8 | 89 | 13.5 | |
| Estrogen receptor (ER) status | |||||||||
| Positive | 632 | 80.3 | 279 | 74.8 | 0.03 | ||||
| Negative | 155 | 19.7 | 94 | 25.2 | |||||
| Progesterone receptor (PR) status | |||||||||
| Positive | 518 | 69.1 | 242 | 67.6 | 0.62 | ||||
| Negative | 232 | 30.9 | 116 | 32.4 | |||||
| ER/PR status | |||||||||
| ER+/PR+ | 506 | 67.6 | 228 | 64.0 | 0.04 | ||||
| ER+/PR− | 90 | 12.0 | 35 | 9.8 | |||||
| ER−/PR+ | 11 | 1.5 | 13 | 3.7 | |||||
| ER−/PR− | 141 | 18.9 | 80 | 22.5 | |||||
| TNRC9 (rs3803662 G>A)b | 1173 | 30 | 1328 | 27 | 564 | 40 | 714 | 39 | <0.001 |
| FGFR2 (rs1219648 A>G)b | 1172 | 43 | 1328 | 40 | 565 | 47 | 714 | 41 | 0.56 |
| FGFR2 (rs2981582 G>A)b | 1172 | 42 | 1326 | 40 | 562 | 46 | 714 | 41 | 0.31 |
| MAP3K1 (rs889312 A>C)b | 1171 | 29 | 1327 | 29 | 564 | 44 | 714 | 41 | <0.001 |
| 2q35 (rs13387042 G>A)b,c | 1169 | 54 | 1328 | 52 | 564 | 41 | 713 | 36 | <0.001 |
Chi-square test of across race/ethnicity differences; within controls only for genotypes
Minor allele frequency
Minor allele assigned for NHW based on minor allele (A) for Hispanic
No significant ethnic differences in associations were detected between NHW and Hispanic women for any of the five SNPs assessed, although magnitude of the association varied for NHW and Hispanic women (Table 2). TNRC9 rs3803662 AA genotype was associated with increased risk among NHW women (OR 1.54, 95% CI 1.14, 2.08; P trend 0.003) but not among Hispanic women (OR 1.05, 95% CI 0.75, 1.46; P trend 0.82). Both polymorphisms of FGFR2 were associated with statistically significant increased risk for both NHW and Hispanic women, while MAP3K1 was not associated with risk among either ethnic group. The polymorphism on 2q35 (rs13387042) was associated with a statistically significant increased risk only among Hispanic women (OR 1.53, 95% CI 1.08, 2.15 for the AA genotype; P trend = 0.004).
Table 2.
Associations between GWAS-identified SNPs and breast cancer in NHW and Hispanic women living in the Southwestern US
| White, non-Hispanic | Hispanic | P heterogeneityb | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Controls | Cases | Controls | Cases | ||||||
| n | n | OR | (95% CI) | n | n | ORa | (95% CI) | ||
| TNRC9 (rs3803662) | 0.11 | ||||||||
| GG | 708 | 569 | 1.00 | 270 | 209 | 1.00 | |||
| GA | 530 | 495 | 1.17 | (0.99, 1.38) | 332 | 260 | 1.00 | (0.78, 1.28) | |
| AA | 90 | 109 | 1.54 | (1.14, 2.08) | 112 | 95 | 1.05 | (0.75, 1.46) | |
| P trend | 0.003 | 0.82 | |||||||
| FGFR2 (rs1219648) | |||||||||
| AA | 474 | 380 | 1.00 | 253 | 150 | 1.00 | |||
| AG | 645 | 582 | 1.12 | (0.94, 1.33) | 337 | 297 | 1.44 | (1.12, 1.87) | |
| GG | 209 | 210 | 1.26 | (1.00, 1.59) | 124 | 118 | 1.53 | (1.10, 2.12) | |
| P trend | 0.05 | 0.005 | |||||||
| FGFR2 (rs2981582) | 0.50 | ||||||||
| GG | 487 | 384 | 1.00 | 254 | 151 | 1.00 | |||
| GA | 642 | 581 | 1.14 | (0.96, 1.36) | 339 | 303 | 1.48 | (1.14, 1.91) | |
| AA | 197 | 207 | 1.33 | (1.05, 1.68) | 121 | 108 | 1.44 | (1.03, 2.00) | |
| P trend | 0.02 | 0.01 | |||||||
| MAP3K1 (rs889312) | 0.33 | ||||||||
| AA | 673 | 587 | 1.00 | 249 | 181 | 1.00 | |||
| AC | 528 | 481 | 1.05 | (0.89, 1.24) | 342 | 273 | 1.13 | (0.87, 1.45) | |
| CC | 126 | 103 | 0.93 | (0.70, 1.24) | 123 | 110 | 1.19 | (0.86, 1.66) | |
| P trend | 0.98 | 0.25 | |||||||
| 2q35 (rs13387042) | 0.10 | ||||||||
| GG | 302 | 257 | 1.00 | 299 | 191 | 1.00 | |||
| GA | 659 | 562 | 1.00 | (0.82, 1.22) | 315 | 278 | 1.39 | (1.09, 1.78) | |
| AA | 367 | 350 | 1.11 | (0.89, 1.39) | 99 | 95 | 1.53 | (1.08, 2.15) | |
| P trend | 0.32 | 0.004 | |||||||
Odds ratios adjusted for age and center and genetic admixture among Hispanic women
P value for heterogeneity test difference between NHW and Hispanic women
No meaningful differences were detected for breast cancer risk between NHW and Hispanic women who were post-menopausal (Table 3). However, associations were significantly different among pre/peri-menopausal women for TNRC9 rs3803662 (P heterogeneity 0.008) and of borderline significant difference for 2q35 rs13387042 (P heterogeneity 0.08) for NHW and Hispanic women. The AA genotype of TNRC9 was associated with over a twofold increased risk of breast cancer among pre/peri-menopausal NHW women (OR 2.18, 95% CI 1.28, 3.73; P linear trend 0.0003) and was not associated with risk among Hispanic women (OR 0.86, 95% CI 0.51, 1.44; P linear trend 0.67). On the other hand the AA genotype of 2q35 rs13387042 significantly increased risk among Hispanic women (OR 1.94, 95% CI 1.12, 3.36; P linear trend 0.01) but was not associated with breast cancer among NHW pre/peri-menopausal women (OR 1.02, 95% CI 0.70, 1.50; P linear trend 0.88).
Table 3.
Associations between GWAS-identified loci and breast cancer among pre/peri and post-menopausal women living in the Southwestern US
| White, non-Hispanic | Hispanics | P heterogeneityb | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Controls | Cases | Controls | Cases | ||||||
| n | n | OR | (95% CI) | n | n | ORa | (95% CI) | ||
| Pre/peri-menopausal | |||||||||
| TNRC9 (rs3803662) | |||||||||
| GG | 236 | 188 | 1.00 | 96 | 85 | 1.00 | 0.008 | ||
| GA | 151 | 182 | 1.51 | (1.13, 2.02) | 114 | 105 | 1.06 | (0.71, 1.59) | |
| AA | 25 | 43 | 2.18 | (1.28, 3.73) | 50 | 40 | 0.86 | (0.51, 1.44) | |
| P trend | 0.0003 | 0.67 | |||||||
| FGFR2 (rs1219648) | |||||||||
| AA | 146 | 128 | 1.00 | 96 | 62 | 1.00 | 0.56 | ||
| AG | 204 | 219 | 1.21 | (0.89, 1.64) | 117 | 123 | 1.70 | (1.12, 2.58) | |
| GG | 62 | 67 | 1.23 | (0.81, 1.88) | 47 | 45 | 1.42 | (0.84, 2.40) | |
| P trend | 0.25 | 0.10 | |||||||
| FGFR2 (rs2981582) | |||||||||
| GG | 151 | 130 | 1.00 | 93 | 60 | 1.00 | 0.76 | ||
| GA | 203 | 217 | 1.24 | (0.91, 1.68) | 118 | 124 | 1.72 | (1.13, 2.62) | |
| AA | 58 | 66 | 1.32 | (0.86, 2.03) | 48 | 45 | 1.39 | (0.82, 2.35) | |
| P trend | 0.14 | 0.12 | |||||||
| MAP3K1 (rs889312) | |||||||||
| AA | 198 | 199 | 1.00 | 80 | 71 | 1.00 | 0.71 | ||
| AC | 163 | 174 | 1.06 | (0.79, 1.42) | 131 | 116 | 1.08 | (0.72, 1.64) | |
| CC | 51 | 40 | 0.81 | (0.51, 1.28) | 48 | 43 | 1.00 | (0.59, 1.71) | |
| P trend | 0.62 | 0.93 | |||||||
| 2q35 (rs13387042) | |||||||||
| GG | 93 | 91 | 1.00 | 113 | 80 | 1.00 | 0.08 | ||
| GA | 195 | 190 | 0.96 | (0.67, 1.37) | 114 | 107 | 1.36 | (0.91, 2.03) | |
| AA | 124 | 131 | 1.02 | (0.70, 1.50) | 33 | 43 | 1.94 | (1.12, 3.36) | |
| P trend | 0.88 | 0.01 | |||||||
| Post-menopausal | |||||||||
| TNRC9 (rs3803662) | |||||||||
| GG | 472 | 379 | 1.00 | 173 | 123 | 1.00 | 0.89 | ||
| GA | 378 | 312 | 1.04 | (0.85, 1.28) | 217 | 154 | 0.98 | (0.72, 1.34) | |
| AA | 65 | 66 | 1.29 | (0.89, 1.87) | 62 | 55 | 1.19 | (0.77, 1.85) | |
| P trend | 0.25 | 0.54 | |||||||
| FGFR2 (rs1219648) | |||||||||
| AA | 327 | 251 | 1.00 | 157 | 88 | 1.00 | 0.34 | ||
| AG | 441 | 361 | 1.07 | (0.86, 1.33) | 219 | 172 | 1.32 | (0.95, 1.84) | |
| GG | 147 | 143 | 1.29 | (0.97, 1.72) | 76 | 73 | 1.60 | (1.05, 2.43) | |
| P trend | 0.09 | 0.02 | |||||||
| FGFR2 (rs2981582) | |||||||||
| GG | 335 | 253 | 1.00 | 161 | 91 | 1.00 | 0.60 | ||
| GA | 439 | 362 | 1.10 | (0.88, 1.36) | 220 | 177 | 1.35 | (0.97, 1.88) | |
| AA | 139 | 141 | 1.35 | (1.01, 1.80) | 72 | 63 | 1.46 | (0.95, 2.25) | |
| P trend | 0.05 | 0.06 | |||||||
| MAP3K1 (rs889312) | |||||||||
| AA | 475 | 386 | 1.00 | 169 | 110 | 1.00 | 0.29 | ||
| AC | 365 | 307 | 1.04 | (0.84, 1.27) | 211 | 156 | 1.15 | (0.83, 1.58) | |
| CC | 74 | 62 | 1.02 | (0.71, 1.47) | 73 | 66 | 1.36 | (0.89, 2.07) | |
| P trend | 0.78 | 0.15 | |||||||
| 2q35 (rs13387042) | |||||||||
| GG | 209 | 165 | 1.00 | 184 | 109 | 1.00 | 0.40 | ||
| GA | 463 | 370 | 1.00 | (0.78, 1.28) | 201 | 171 | 1.44 | (1.05, 1.98) | |
| AA | 243 | 219 | 1.14 | (0.86, 1.49) | 66 | 52 | 1.32 | (0.85, 2.05) | |
| P trend | 0.34 | 0.08 | |||||||
Odds ratios (OR) adjusted for age and center and genetic admixture among Hispanic women
P value for heterogeneity test difference between NHW and Hispanic women
Associations by ER/PR tumor status (Table 4) revealed few findings. Among NHW women only 2q35 rs13387042 appeared to be uniquely associated with ER+/PR+ tumors. Both FGFR2 polymorphisms reduced risk of ER−PR− tumors in the presence of the minor allele among NHW women. Among Hispanic women, both polymorphisms of the FGFR2 gene were associated with almost a twofold increase risk of an ER+/PR+ tumor in a dose–response manner, while non-significantly inversely associated with ER−/PR− tumors.
Table 4.
Associations between GWAS loci and ER/PR status among breast cancer cases living in the Southwestern US
| Controls | ER+/PR+ cases | ER+/PR− cases | ER−/PR−b cases | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| n | n | ORa | (95% CI) | n | OR | (95% CI) | n | OR | (95% CI) | |
| White, non-Hispanic | ||||||||||
| TNRC9 (rs3803662) | ||||||||||
| GG | 708 | 245 | 1.00 | 41 | 1.00 | 69 | 1.00 | |||
| GA | 530 | 209 | 1.13 | (0.91, 1.41) | 42 | 1.33 | (0.85, 2.09) | 58 | 1.15 | (0.79, 1.66) |
| AA | 90 | 48 | 1.58 | (1.08, 2.31) | 8 | 1.56 | (0.71, 3.46) | 14 | 1.71 | (0.92, 3.18) |
| P trend | 0.03 | 0.14 | 0.12 | |||||||
| FGFR2 (rs1219648) | ||||||||||
| AA | 474 | 169 | 1.00 | 27 | 1.00 | 63 | 1.00 | |||
| AG | 645 | 249 | 1.08 | (0.86, 1.35) | 49 | 1.35 | (0.83, 2.19) | 57 | 0.67 | (0.46, 0.98) |
| GG | 209 | 84 | 1.14 | (0.83, 1.55) | 14 | 1.20 | (0.61, 2.34) | 20 | 0.74 | (0.43, 1.26) |
| AG/GG | 854 | 333 | 1.09 | (0.88, 1.36) | 63 | 1.29 | (0.81, 2.06) | 77 | 0.69 | (0.48, 0.98) |
| P trend | 0.39 | 0.43 | 0.10 | |||||||
| FGFR2 (rs2981582) | ||||||||||
| GG | 487 | 174 | 1.00 | 26 | 1.00 | 64 | 1.00 | |||
| GA | 642 | 251 | 1.09 | (0.87, 1.37) | 49 | 1.45 | (0.89, 2.37) | 56 | 0.67 | (0.46, 0.97) |
| AA | 197 | 78 | 1.10 | (0.81, 1.51) | 15 | 1.46 | (0.75, 2.82) | 19 | 0.75 | (0.44, 1.29) |
| GA/AA | 839 | 329 | 1.10 | (0.88, 1.36) | 64 | 1.43 | (0.89, 2.29) | 75 | 0.69 | (0.48, 0.98) |
| P trend | 0.46 | 0.17 | 0.10 | |||||||
| MAP3K1 (rs889312) | ||||||||||
| AA | 673 | 259 | 1.00 | 49 | 1.00 | 70 | 1.00 | |||
| AC | 528 | 208 | 1.03 | (0.83, 1.28) | 31 | 0.80 | (0.50, 1.28) | 60 | 1.10 | (0.76, 1.58) |
| CC | 126 | 34 | 0.70 | (0.47, 1.05) | 11 | 1.19 | (0.60, 2.35) | 9 | 0.65 | (0.32, 1.34) |
| P trend | 0.29 | 0.91 | 0.58 | |||||||
| 2q35 (rs13387042) | ||||||||||
| GG | 302 | 98 | 1.00 | 25 | 1.00 | 34 | 1.00 | |||
| GA | 659 | 244 | 1.15 | (0.87, 1.51) | 43 | 0.77 | (0.46, 1.29) | 68 | 0.90 | (0.58, 1.40) |
| AA | 367 | 159 | 1.34 | (1.00, 1.80) | 22 | 0.73 | (0.40, 1.32) | 37 | 0.91 | (0.55, 1.48) |
| P trend | 0.05 | 0.30 | 0.70 | |||||||
| Hispanic | ||||||||||
| TNRC9 (rs3803662) | ||||||||||
| GG | 270 | 82 | 1.00 | 10 | 1.00 | 35 | 1.00 | |||
| GA | 332 | 105 | 1.02 | (0.73, 1.43) | 17 | 1.40 | (0.62, 3.14) | 36 | 0.83 | (0.50, 1.36) |
| AA | 112 | 40 | 1.10 | (0.71, 1.72) | 8 | 2.21 | (0.83, 5.85) | 9 | 0.54 | (0.25, 1.17) |
| P trend | 0.69 | 0.12 | 0.12 | |||||||
| FGFR2 (rs1219648) | ||||||||||
| AA | 253 | 53 | 1.00 | 10 | 1.00 | 30 | 1.00 | |||
| AG | 337 | 122 | 1.69 | (1.18, 2.43) | 20 | 1.51 | (0.67, 3.24) | 38 | 0.90 | (0.54, 1.50) |
| GG | 124 | 53 | 1.95 | (1.26, 3.03) | 5 | 0.97 | (0.33, 3.05) | 12 | 0.76 | (0.37, 1.55) |
| 461 | 175 | 1.76 | (1.25, 2.49) | 25 | 1.36 | (0.63, 2.89) | 50 | 0.86 | (0.53, 1.40) | |
| P trend | 0.002 | 0.83 | 0.45 | |||||||
| FGFR2 (rs2981582) | ||||||||||
| GG | 254 | 55 | 1.00 | 10 | 1.00 | 29 | 1.00 | |||
| GA | 339 | 118 | 1.58 | (1.10, 2.27) | 20 | 1.47 | (0.67, 3.24) | 40 | 0.99 | (0.60, 1.66) |
| AA | 121 | 54 | 1.97 | (1.27, 3.05) | 5 | 1.00 | (0.33, 3.05) | 11 | 0.73 | (0.35, 1.52) |
| 460 | 172 | 1.69 | (1.20, 2.38) | 25 | 1.35 | (0.63, 2.89) | 51 | 0.92 | (0.57, 1.51) | |
| P trend | 0.002 | 0.79 | 0.47 | |||||||
| MAP3K1 (rs889312) | ||||||||||
| AA | 249 | 74 | 1.00 | 13 | 1.00 | 23 | 1.00 | |||
| AC | 342 | 113 | 1.12 | (0.80, 1.57) | 14 | 0.85 | (0.39, 1.85) | 45 | 1.41 | (0.83, 2.41) |
| CC | 123 | 40 | 1.04 | (0.67, 1.63) | 8 | 1.27 | (0.50, 3.19) | 12 | 0.93 | (0.45, 1.96) |
| P trend | 0.75 | 0.73 | 0.87 | |||||||
| 2q35 (rs13387042) | ||||||||||
| GG | 299 | 69 | 1.00 | 11 | 1.00 | 25 | 1.00 | |||
| GA | 315 | 120 | 1.71 | (1.21, 2.40) | 19 | 1.53 | (0.71, 3.32) | 41 | 1.57 | (0.92, 2.66) |
| AA | 99 | 38 | 1.75 | (1.10, 2.79) | 5 | 1.53 | (0.51, 4.57) | 14 | 1.60 | (0.79, 3.25) |
| P trend | 0.004 | 0.32 | 0.12 | |||||||
Odds ratios adjusted for age and center and genetic admixture among Hispanic women
ER−/PR+ cases were too few to analyze
Discussion
Our replication of five SNPs identified from multiple GWAS of breast cancer in Hispanic and NHW women living in the Southwestern United States, showed that three SNPs replicated in Hispanic women, FGFR2 rs1219648 and rs2981582 and the SNP located on 2q35 (rs13387042). Three SNPs also replicated among NHW women, TNRC9 rs3803662 and both SNPs on FGFR2. While associations did not differ by family history, they appear to be slightly stronger for pre/peri-menopausal women. Of interest is the association between both FGFR2 polymorphisms and ER/PR tumor status, in that among NHW women the minor allele is associated with reduced risk of0 ER−/PR− tumors. Among Hispanic women, there is a dose response increase in risk of an ER+/PR+ tumor associated with the minor allele of the FGFR2 polymorphisms.
The two polymorphisms in FGFR2 are in high LD with an r2 of 0.93 for NHW and 0.88 for Hispanic women. Thus it is not surprising that findings for the two polymorphisms are similar. These associations for these SNPs were first reported by Easton and Hunter [1, 2] have since been replicated in several populations including studies of African Americans [12]. Hunter’s report of a pooled estimate of a 20% increased risk with the heterozygote and 64% increased risk for homozygote variant among post-menopausal women is higher than we report here, although a similar trend in risk is observed. Easton reported findings similar to those by Hunter. Our risk estimates for post-menopausal Hispanic women are more similar to those reported by Hunter and Easton than are our findings for NHW women. FGFR2 is a tumor suppressor gene that is often over-expressed in breast tumors [13]. It has been hypothesized that FGFR2 is involved in signal transduction and transformation of mammary epithelial cell lines.
The findings for TNRC9, MAP3K1, and 2q35 have been replicated in some, but not other studies. TNRC9 rs3803662 is near the 5′ end of TNRC9, whose protein expression is implicated in breast cancer metastasis to bone [7]. The association with TNRC9 rs3803662 was not confirmed in two studies of African American women [6, 12] although it has been shown to have strong signals in other studies, including studies involving deep sequencing of the gene [3, 8]. Neither MAP3K1 rs889312 nor rs13387042 on 2q35 have been confirmed in studies of African American women [6, 12], however, another polymorphism of MAP3K1, rs16886165, has been reported in a multi-stage breast cancer GWAS study [3].
Several studies have examined these loci with breast cancer clinical features, including ER/PR status of tumors. A study by Garcia-Closas and Chanock [14] did not observe any associations between these loci and ER− tumors. The increased risk was restricted to ER+ tumors and was seen for all five SNPs. Nordgard and colleagues observed TNRC9 was up-regulated in luminal A, luminal B, and ErbB2+ tumors and down-regulated in basal-like subtype [15]. Our data suggest that FGFR2 may influence ER/PR status of tumors, decreasing the likelihood of ER− tumors among NHW women while increasing the likelihood of ER/PR+ tumors among Hispanic women. Evaluation of molecular subtypes of breast cancer with GWAS data has been cited as an important component to identification of susceptibility loci [16].
Conclusions
Our results support the hypothesis that genetic factors differ by race and ethnicity as they relate to breast cancer. The importance of examining ethnicity/race associations as a component of validating and replicating associations is critical to understand the complexity of disease associations among genetically admixed populations. Because of these differences, it also is important to examine GWAS in Hispanic admixed populations to identify polymorphisms that may be uniquely associated with these populations and could account for biological disparities in breast cancer incidence.
Acknowledgments
CA 078682, CA 078762, CA078552, CA078802, and CA14002. This research also was supported by the Utah Cancer Registry, which is funded by Contract #N01-PC-67000 from the National Cancer Institute, with additional support from the State of Utah Department of Health, the New Mexico Tumor Registry, funded by contract #, and the Arizona and Colorado cancer registries, funded by the Centers for Disease Control and Prevention National Program of Cancer Registries and additional state support. The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official view of the National Cancer Institute. We would like to acknowledge the contributions of Dr. Betsy Risendal, Sandra Edwards, Roger Edwards, Leslie Palmer, Tara Patton, Jason Witter, and Kelly May to this study.
Abbreviations
- NHW
Non-Hispanic white
- OR
Odds ratios
- CI
Confidence intervals
- ER
Estrogen receptor
- PR
Progesterone receptor
- GWAS
Genome-wide association study
- MAP
Mitogen-activated protein
- FGFR
Fibroblast growth factor receptor
- TRNC9
Trinucleotide repeat-containing 9
- GUESS
Generally useful ethnic search system
- AI
American Indian
- SNP
Single nucleotide polymorphism
- HWE
Hardy–Weinberg Equilibrium
- MAF
Minor allele frequency
Footnotes
Conflict of interest None.
Contributor Information
Martha L. Slattery, University of Utah, Salt Lake City, UT 84108, USA, marty.slattery@hsc.utah.edu
Kathy B. Baumgartner, Department of Epidemiology and Population Health, School of Public Health and Information Sciences, University of Louisville, Louisville, KY, USA
Anna R. Giuliano, Moffitt Cancer Center, Tampa, FL, USA
Tim Byers, University of Colorado School of Medicine, Denver, CO, USA.
Jennifer S. Herrick, University of Utah, Salt Lake City, UT 84108, USA
Roger K. Wolff, University of Utah, Salt Lake City, UT 84108, USA
References
- 1.Easton DF, Pooley KA, Dunning AM, Pharoah PD, Thompson D, Ballinger DG, Struewing JP, Morrison J, Field H, Luben R, et al. Genome-wide association study identifies novel breast cancer susceptibility loci. Nature. 2007;447(7148):1087–1093. doi: 10.1038/nature05887. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Hunter DJ, Kraft P, Jacobs KB, Cox DG, Yeager M, Hankinson SE, Wacholder S, Wang Z, Welch R, Hutchinson A, et al. A genome-wide association study identifies alleles in FGFR2 associated with risk of sporadic postmenopausal breast cancer. Nat Genet. 2007;39(7):870–874. doi: 10.1038/ng2075. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Thomas G, Jacobs KB, Kraft P, Yeager M, Wacholder S, Cox DG, Hankinson SE, Hutchinson A, Wang Z, Yu K, et al. A multistage genome-wide association study in breast cancer identifies two new risk alleles at 1p11.2 and 14q24.1 (RAD51L1) Nat Genet. 2009;41(5):579–584. doi: 10.1038/ng.353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ruiz-Narvaez EA, Rosenberg L, Cozier YC, Cupples LA, Adams-Campbell LL, Palmer JR. Polymorphisms in the TOX3/LOC643714 locus and risk of breast cancer in African-American women. Cancer Epidemiol Biomarkers Prev. 2010;19(5):1320–1327. doi: 10.1158/1055-9965.EPI-09-1250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Long J, Shu XO, Cai Q, Gao YT, Zheng Y, Li G, Li C, Gu K, Wen W, Xiang YB, et al. Evaluation of breast cancer susceptibility loci in Chinese women. Cancer Epidemiol Biomarkers Prev. 2010;19(9):2357–2365. doi: 10.1158/1055-9965.EPI-10-0054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Zheng W, Cai Q, Signorello LB, Long J, Hargreaves MK, Deming SL, Li G, Li C, Cui Y, Blot WJ. Evaluation of 11 breast cancer susceptibility loci in African-American women. Cancer Epidemiol Biomarkers Prev. 2009;18(10):2761–2764. doi: 10.1158/1055-9965.EPI-09-0624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Stacey SN, Manolescu A, Sulem P, Rafnar T, Gudmundsson J, Gudjonsson SA, Masson G, Jakobsdottir M, Thorlacius S, Helgason A, et al. Common variants on chromosomes 2q35 and 16q12 confer susceptibility to estrogen receptor-positive breast cancer. Nat Genet. 2007;39(7):865–869. doi: 10.1038/ng2064. [DOI] [PubMed] [Google Scholar]
- 8.Udler MS, Ahmed S, Healey CS, Meyer K, Struewing J, Maranian M, Kwon EM, Zhang J, Tyrer J, Karlins E, et al. Fine scale mapping of the breast cancer 16q12 locus. Hum Mol Genet. 2010;19(12):2507–2515. doi: 10.1093/hmg/ddq122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Slattery ML, Sweeney C, Edwards S, Herrick J, Baumgartner K, Wolff R, Murtaugh M, Baumgartner R, Giuliano A, Byers T. Body size, weight change, fat distribution and breast cancer risk in Hispanic and non-Hispanic white women. Breast Cancer Res Treat. 2007;102(1):85–101. doi: 10.1007/s10549-006-9292-y. [DOI] [PubMed] [Google Scholar]
- 10.Howard CA, Samet JM, Buechley RW, Schrag SD, Key CR. Survey research in New Mexico Hispanics: some methodological issues. Am J Epidemiol. 1983;117(1):27–34. doi: 10.1093/oxfordjournals.aje.a113512. [DOI] [PubMed] [Google Scholar]
- 11.Sweeney C, Wolff RK, Byers T, Baumgartner KB, Giuliano AR, Herrick JS, Murtaugh MA, Samowitz WS, Slattery ML. Genetic admixture among Hispanics and candidate gene polymorphisms: potential for confounding in a breast cancer study? Cancer Epidemiol Biomarkers Prev. 2007;16(1):142–150. doi: 10.1158/1055-9965.EPI-06-0706. [DOI] [PubMed] [Google Scholar]
- 12.Barnholtz-Sloan JS, Shetty PB, Guan X, Nyante SJ, Luo J, Brennan DJ, Millikan RC. FGFR2 and other loci identified in genome-wide association studies are associated with breast cancer in African-American and younger women. Carcinogenesis. 2010;31(8):1417–1423. doi: 10.1093/carcin/bgq128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Grose R, Dickson C. Fibroblast growth factor signaling in tumorigenesis. Cytokine Growth Factor Rev. 2005;16(2):179–186. doi: 10.1016/j.cytogfr.2005.01.003. [DOI] [PubMed] [Google Scholar]
- 14.Garcia-Closas M, Hall P, Nevanlinna H, Pooley K, Morrison J, Richesson DA, Bojesen SE, Nordestgaard BG, Axelsson CK, Arias JI, et al. Heterogeneity of breast cancer associations with five susceptibility loci by clinical and pathological characteristics. PLoS Genet. 2008;4(4):e1000054. doi: 10.1371/journal.pgen.1000054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Nordgard SH, Johansen FE, Alnaes GI, Naume B, Borresen-Dale AL, Kristensen VN. Genes harbouring susceptibility SNPs are differentially expressed in the breast cancer subtypes. Breast Cancer Res. 2007;9(6):113. doi: 10.1186/bcr1784. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kristensen VN, Borresen-Dale AL. SNPs associated with molecular subtypes of breast cancer: on the usefulness of stratified Genome-wide Association Studies (GWAS) in the identification of novel susceptibility loci. Mol Oncol. 2008;2(1):12–15. doi: 10.1016/j.molonc.2008.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]