Abstract
Background
In addition to mediating aspects of physiological and pathological angiogenesis, the vascular endothelial growth factor (VEGF) family also contributes to carcinogenesis.
Methods
We comprehensively characterized genetic variation across four VEGF family genes and evaluated associations with breast cancer risk with odds ratios and 95% confidence intervals (OR, 95% CI) for participants of the two-stage case-control Shanghai Breast Cancer Genetics Study (SBCGS). Stage 1 evaluated 200 single nucleotide polymorphisms (SNPs) across two VEGF ligands (VEGFA and VEGFC) and two VEGF receptors (FLT1/VEGFR1 and KDR/VEGFR2) among 2,079 cases and 2,148 controls. Five SNPs with promising associations were assessed in Stage 2 among 4,419 cases and 1,851 controls.
Results
Two SNPs were consistently associated with breast cancer risk across our two study stages, and significant in combined analyses. Compared to FLT1 rs9551471 major allele homozygotes (AA), reduced risks were associated with AG (OR=0.92, 95% CI: 0.84-1.00) and GG (OR=0.78, 95% CI: 0.64-0.95) genotypes (p-value for trend = 0.005). Compared to VEGFA rs833070 major allele carriers (CC or CT), increased risk was associated with TT genotypes (OR=1.26, 95% CI: 1.05-1.52, p-value = 0.016).
Conclusion
Results from our study indicate that common genetic variation in VEGFA and FLT1 (VEGFR1) may contribute to breast cancer susceptibility.
Impact
Our findings provide clues for future studies on VEGF family genes in relation to cancer susceptibility and survival.
Keywords: VEGF, VEGFA, FLT1, polymorphisms, breast cancer, susceptibility
Introduction
The human vascular endothelial growth factor (VEGF) family includes five growth factors and three tyrosine kinase receptors that mediate various aspects of both physiological and pathological angiogenesis, lymphangiogenesis, and vasculogenesis (1-3). In addition to normal processes, such as ovarian follicular development and embryogenesis, VEGF family members are also involved in several diseases, including macular degeneration, rheumatoid arthritis, endometriosis, and cardiovascular disease (2-6). In regards to cancer, the actions of the VEGF family are crucial for tumor growth and metastasis, as tumors cannot grow beyond a limited initial size without the establishment of blood flow for acquiring nutrients and carrying away debris (2, 3). In addition to this angiogenic switch, VEGF family members can also contribute to cancer development by promoting cell growth and enhancing cell survival (1, 2, 4, 6, 7).
The members of the VEGF family have both unique and overlapping functions, which are mediated by specific ligand-receptor interactions. VEGFA is a potent endothelial mitogen, that also regulates vascular permeability, cellular motility, and survival (2-4, 7). VEGFC influences the migration and survival of lymphatic endothelial cells and supports the maintenance of differentiated lymphatic endothelium (2, 3). Other VEGF family growth factors include placental growth factor (PGF), which enhances VEGF signaling, FIGF/VEGFD (c-fos induced growth factor), which contributes to lymphangiogenesis, and VEGFB, the function of which remains unclear (2, 3). Due to alternative splicing, multiple isoforms of VEGFA, VEGFB, and PlGF exist; these isoforms differ in their mRNA stability, amino acid sequences, protein localization, and functions (1-3). The vascular endothelial growth factor receptor 1, known as Fms-like tyrosine kinase-1 (FLT1), has highest affinity for VEGFA, but can also bind VEGFB or PlGF (1, 2, 8). As VEGFA binding results in only a small increase in kinase activity, FLT1 functions primarily as a negative regulator of angiogenesis by sequestering VEGFA; however, positive roles in the regulation of vascular permeability, cell motility, and apoptotic inhibition have also been identified (1-3, 8). FLT1 also undergoes alternative splicing to encode either a membrane bound receptor or a soluble variant (sFLT1) that acts to bind VEGFA and block its signaling (1-3). VEGFR2, also known as KDR (Kinase-insert Domain containing Receptor), has lower affinity for VEGFA than FLT1, but a much higher protein-tyrosine kinase activity (1-3). KDR is the primary mediator of normal angiogeneic signaling by VEGFA, but can also bind VEGFC and FIGF (1-3). VEGFR3, also known as FLT4, binds to VEGFC and FIGF, and contributes to angiogenesis, lymphangiogenesis, and cell survival (1, 2).
Thus, VEGFA is the key mediator of angiogenesis, and therefore intertwined in cancer development via the angiogenic switch (1-4). Further, VEGFA and other VEGF family genes also have crucial roles in cell survival, mitogenesis, migration, differentiation, vascular permeability, and mobilization, and therefore may influence cancer development (1, 2, 4, 8). Common genetic variation in VEGF family genes may influence gene transcription, mRNA stability, protein splicing, and protein functions, and therefore cancer risk. Several previous studies have examined genetic variation in 8 single nucleotide polymorphisms (SNPs) in VEGF-A (rs699947, rs1005230, rs833061, rs1570360, rs2010963, rs25648, rs10434, and rs3025039) in relation to breast cancer risk; results have been inconsistent (9-18). Studies of genetic variation in other VEGF family genes and breast cancer risk are sparse; one FLT1 SNP (-962 C/T) and two KDR SNPs (rs2305948 and rs1870377) were evaluated in two studies with null results (18, 19). Further, no studies employed a two-stage genotyping approach to minimize false-positive findings. We therefore conducted a large, two-stage study, in which we comprehensively evaluated genetic variation in four major genes in the VEGF family in relation to breast cancer risk among participants of the Shanghai Breast Cancer Genetics Study.
Materials and Methods
Study Population
Subjects were participants of three population based studies conducted among Chinese women in urban Shanghai: the Shanghai Breast Cancer Study (SBCS), the Shanghai Breast Cancer Survival Study (SBCSS), and the Shanghai Endometrial Cancer Study (SECS). Collectively, these studies comprise the Shanghai Breast Cancer Genetics Study (SBCGS); detailed methodologies for these studies have been previously reported (20). Briefly, the SBCS is a large two-stage (SBCS-I and SBCS-II), population-based, case-control study. Breast cancer cases were identified via a rapid case-ascertainment system and the Shanghai Cancer Registry; diagnoses were confirmed by two senior pathologists. Controls were randomly selected using the Shanghai Resident Registry. SBCS-I recruitment occurred between August 1996 and March 1998, and included women aged 25-65. SBCS-II recruitment occurred from April 2002 to February 2005, and was expanded to include women aged 20-70. Also included in the study were cases recruited between April 2002 and December 2006 as part of the SBCSS, a population-based study of newly diagnosed breast cancer cases identified by the Shanghai Cancer Registry, and controls recruited between January 1997 and December 2003 as part of the SECS, a population-based case-control study of endometrial cancer that included women aged 25-70 that was conducted in a manner almost identical to the SBCS. Of those eligible, 1,459 cases (91.1%) and 1,556 controls (90.3%) from SBCS-I, 1,989 cases (83.7%) and 1,989 (70.4%) controls from SBCS-II, and 5,046 cases (80.1%) from the SBCSS and 1,212 controls (74.4%) from the SECS completed in person interviews with structured questionnaires. Blood or buccal cell samples were donated and available for 1,193 cases (81.8%) and 1,310 controls (84.2%) from SBCS-I, 1,932 cases (97.1%) and 1,857 controls (93.4%) from SBCS-II, and 4,845 (96.0%) cases from the SBCSS and 1,039 (85.7%) controls from the SECS. Because of a time overlap in subject recruitment, 1,469 breast cancer patients participated in both the SBCS-II and the SBCSS, and 109 controls participated in both the SBCS-I and SECS, so that the actual number of participants included in the current analysis from SBCSS and SECS was 3,466 and 930, respectively. For breast cancer cases, clinical characteristics were ascertained by medical record abstraction using a standard protocol. Stage of disease was available for 5,992 cases (92.2%). Of these breast cancer cases, 36.1% were stage 0 or 1, 34.7% were stage 2, 18.6% were stage 3, and 10.7% were stage 4. Estrogen receptor (ER) status was available for 5,967 cases (91.8%); 64.2% of these cases were ER positive and 36.8% were ER negative. Progesterone receptor (PR) status was available for 5,941 cases (91.4%); 59.5% of these cases were PR positive and 40.6% were PR negative. Genomic DNA for all included participants was extracted using commercial DNA purification kits. Informed consent was granted by all included women, and approval was granted from relevant review boards in both China and the United States.
SNP Selection and Genotyping
Included in the current analysis were SNPs from four VEGF gene family members for which we had high genetic coverage of HapMap SNPs (greater than 80% with a pairwise r2 of 0.8) that had minimum minor allele frequencies (MAF) of 0.05 among Han Chinese (HCB). Stage 1 genotyping included four methods; all available data was utilized to maximize our coverage of genetic variation. First, haplotype tagging SNPs (htSNPs) were selected from Han Chinese data from the HapMap Project using the Tagger program to capture SNPs with a minimum minor allele frequency (MAF) of 0.05 in the VEGFA, VEGFC, FLT1, and KDR genes (± 5kb), with an r2 of 0.90 or greater. Seventy-four htSNPs were successfully designed and genotyped among 2,131 SBCS-I participants using a Targeted Genotyping System (Affymetrix) in 2006. Second, to increase the density of genetic markers in this study, data from our Affymetrix Genome-Wide SNP Array 6.0 (Affymetrix) was included for 150 SNPs that were genotyped among 4,157 SBCS-I and SBCS-II participants. Third, four SNPs (rs833061, rs1570360, rs2010963, and rs3025039) were genotyped by TaqMan among 2,350 SBCS-I participants; rs3025039 was additionally genotyped among 3,777 SBCS-II participants. Finally, two SNPs (rs9554312 and rs9582036) were genotyped with the Sequenom iPLEX MassARRAY platform among 3,777 SBCS-II participants. Twenty-eight SNPs were genotyped by more than one method, generating a total of 200 SNPs. Stage 1 analysis was conducted for 132 SNPs with a minimum MAF of 0.05 among genotyped controls. SNPs with promising associations were further evaluated for linkage disequilibrium (LD), and consistency of associations with breast cancer risk when stratified by SBCS study population, when possible. Stage 2 genotyping among 3,453 SBCSS cases and 914 SECS controls and 966 cases and 937 controls from SBCS that were not genotyped in Stage 1 was conducted with the Sequenom platform as previously described. For all genotyping methods, blinded duplicate samples and quality controls were included. All included SNPs had concordance rates of at least 95% among duplicates within each platform, as well as across genotyping platforms. Specifically, for Affymetrix Targeted genotyping, the average call rate was 99.7%. Blinded (n = 39) and HapMap samples (n = 12) were included, consistency rates averaged 99.6%. For Affymetrix GWAS SNPs analyzed in this study, the average call rate was 99.7%. Three quality control samples were included per plate (20); the average concordance rate was 99.9% for 150 SNPs. For TaqMan targeted genotyping, the average call rate was 96.7. Eight blinded and eight unblinded quality control samples were included per plate; concordance rates were greater than 97% (13). For Sequenom targeted genotyping, the average call rate was 97.7%. Two blinded and two HapMap samples were included per plate; the average concordance rate was 99.5% for seven SNPs. Laboratory personnel were blinded to the case-control status of all samples.
Statistical Analysis
Characteristics between cases and controls were compared with the χ2 test or t-test for categorical or continuous variables, respectively. Stage 1 odds ratios and corresponding confidence intervals (OR, 95% CI) were determined by logistic regression that included adjustment for age and education. For SNPs genotyped in Stage 2, pooled analyses of Stage 1 and Stage 2 data included additional adjustment for genotyping stage. Additive, dominant, and recessive models of effect were employed. Multiplicative interactions between genetic variants and family history of breast cancer or menopausal status were evaluated by comparing estimates of effect across strata for heterogeneity, as well as the statistical significance of interaction terms in logistic regression models. The majority of analyses were conducted with SAS v9.2. Permutation tests were conducted using the max(T) approach in PLINK v1.07 with 1000 iterations to determine the family-wise corrected p-value for SNP-breast cancer associations after correcting for the number of variants evaluated while considering their underlying LD structure (21). All statistical tests were two-tailed, and p-values of ≤ 0.05 were interpreted as statistically significant.
Results
In Stage 1, inherited genetic variation in four VEGF family genes was comprehensively characterized (Table 1). A total of 200 SNPs were genotyped across two VEGF family ligands (VEGFA and VEGFC) and two receptors FLT1 (VEGFR1) and KDR (VEGFR-2); of these, 132 had MAFs of at least 5% among controls. Using a pair-wise tagging approach and an r2 of 0. 9 or 0.8, the coverage of SNPs with minimum MAF of 0.05% identified in Han Chinese (HCB) included in the HapMap project was greater than 77% or 85%, respectively. Twenty-eight SNPs were genotyped by more than one method, the majority of the overlap occurred between Affymetrix targeted genotyping and GWAS data (N=26). Consistency rates ranged from 97.98% to 100% and averaged 99.62%; the average number of participants genotyped by both methods was 1,846.
Table 1. SNP Information for Included VEGF Gene Family Members.
| VEGF Gene Family | Genomic Location | Gene Span (kb) | SNPs in HapMap * | SNPs Genotyped | Genetic Variation Coverage ** | |
|---|---|---|---|---|---|---|
|
| ||||||
| r2 = 0.8 | r2 = 0.9 | |||||
| Ligands | ||||||
| Vascular Endothelial Growth Factor A (VEGFA) | 6p12 | 16.3 | 27 | 22 | 85.2% | 77.8% |
| Vascular Endothelial Growth Factor C (VEGFC) | 4q34 | 109.2 | 59 | 29 | 93.2% | 78.0% |
| Receptors | ||||||
| FMS-Related Tyrosine Kinse 1 (FLT1) / Vascular Endothelial Growth Factor Receptor 1 (VEGFR1) | 13q12 | 193.4 | 127 | 119 | 94.5% | 91.3% |
| Kinase Insert Domain Receptor (KDR) / Vascular Endothelial Growth Factor Receptor 2 (VEGFR2) | 4q12 | 47.1 | 55 | 30 | 98.2% | 96.4% |
HapMap data among Han Chinese (HCB), SNPs with minor allele frequency (MAF) greater than 0.05%
Coverage of HapMap CHB SNPs by our genotyped SNPs, using a pairwise tagging approach in Tagger
A total of 10,497 women were included in the current study; 4,227 women in Stage 1, and 6,270 women in Stage 2. Participants genotyped in the two Stages were generally comparable (Table 2). Established breast cancer risk factors, including early age at menarche, late age at menopause, late age at first live birth, a first degree relative with breast cancer, prior history of fibroadenoma, high body mass index (BMI) or waist-to-hip ratio (WHR), and low physical activity were found to be associated with breast cancer risk among SBCGS participants. Women who did not donate DNA samples (N=1,088) were generally comparable to women who did, although these women were younger, less educated, had earlier ages at menopause, lower BMI and/or WHRs, and were also less likely to engage in regular physical activity than women who did donate DNA samples.
Table 2. Demographic and Other Breast Cancer Risk Factors, By Genotyping Stage, the Shanghai Breast Cancer Genetics Study.
| Characteristics | Study Stage 1 (4,227) | Study Stage 2 (N=6,270) | By Genotyping Status (N=11,585) | ||||||
|---|---|---|---|---|---|---|---|---|---|
|
|
|
|
|||||||
| Cases (N=2,079) | Controls (N=2,148) | p-value | Cases (N=4,419) | Controls (N=1,851) | p-value | Genotyped (N=10,497) | Not Genotyped (N=1,088) | p-value | |
| Demographic Factors | |||||||||
| Age (years) | 49.3 ± 8.3 | 49.1 ± 8.7 | 0.390 | 53.9 ± 10.2 | 53.3 ± 8.8 | 0.016 | 51.8 ± 9.5 | 50.6 ± 9.1 | <0.001 |
| Education (less than middle school) | 206 (9.9%) | 286 (13.3%) | <0.001 | 525 (11.9%) | 304 (16.4%) | <0.001 | 1,255 (12.0%) | 163 (15.0%) | 0.004 |
| Reproductive Risk Factors | |||||||||
| Age at menarche (years) | 14.5 ± 1.7 | 14.7 ± 1.8 | <0.001 | 14.4 ± 1.7 | 14.7 ±1.8 | <0.001 | 14.5 ± 1.7 | 14.5 ±1.7 | 0.902 |
| Postmenopausal | 804 (38.8%) | 873 (40.7%) | 0.195 | 2,257 (51.1%) | 1,047 (56.6%) | <0.001 | 4,988 (47.5%) | 507 (46.6%) | 0.572 |
| Age at menopause (years) 1 | 48.4 ± 4.5 | 47.7 ± 4.8 | 0.002 | 49.0 ± 4.3 | 48.8 ± 4.1 | 0.162 | 48.6 ± 4.4 | 48.2 ± 4.5 | 0.042 |
| Age at first live birth (years) 2 | 26.5 ± 3.9 | 26.0 ± 3.8 | 0.001 | 26.8 ± 3.9 | 25.5 ± 4.0 | <0.001 | 26.8 ± 3.8 | 26.6 ± 4.2 | 0.158 |
| Use of oral contraceptives 3 | 419 (20.2%) | 433 (20.2%) | 0.997 | 166 (17.2%) | 416 (22.5%) | 0.001 | 1,434 (20.4%) | 205 (20.9%) | 0.696 |
| Additional Risk Factors | |||||||||
| First degree relative with breast cancer | 92 (4.4%) | 63 (2.9%) | 0.010 | 246 (5.6%) | 24 (2.6%) | <0.001 | 447 (4.3%) | 37 (3.4%) | 0.182 |
| Use of hormone replacement therapy 4 | 65 (3.1%) | 51 (2.4%) | 0.136 | 161 (5.7%) | 70 (3.8%) | 0.003 | 347 (3.9%) | 26 (2.7%) | 0.054 |
| History of breast fibroadenomas 5 | 201 (9.7%) | 121 (5.6%) | <0.001 | 96 (10.0%) | 48 (5.1%) | <0.001 | 514 (7.3%) | 67 (6.9%) | 0.614 |
| Body mass index (kg/m2) | 23.8 ± 3.3 | 23.3 ± 3.4 | <0.001 | 24.0 ± 3.5 | 23.6 ± 3.3 | <0.001 | 23.7 ± 3.4 | 23.3 ± 3.4 | <0.001 |
| Waist-to-hip ratio | 0.822 ± 0.06 | 0.808 ± 0.06 | <0.001 | 0.834 ± 0.05 | 0.816 ± 0.05 | <0.001 | 0.823 ± 0.06 | 0.812 ± 0.05 | <0.001 |
| Regular physical activity | 521 (25.1%) | 638 (29.7%) | <0.001 | 2,473 (56.0%) | 622 (33.6%) | <0.001 | 4,245 (40.5%) | 299 (27.5%) | <0.001 |
Continuous variables: mean values ± standard deviation, p-value from t-tests; Categorical variables: numbers and percentages, p-values from χ2 test
Among postmenopausal women
Among parous women
Information unavailable for 3,456 participants genotyped in Stage 2
Information unavailable for 1,601 participants genotyped in Stage 2
Information unavailable for 4,375 participants genotyped in Stage 2
Bold values considered to be significant p≤0.05
Associations with breast cancer risk for 132 SNPs with MAFs ≥ 0.05 genotyped in Stage 1 yielded significant p-values (≤0.05) in either additive, dominant, or recessive models for 12 SNPs (Supplemental Table 1). Linkage disequilibrium (LD) and consistency of associations between SBCS-I and SBCS-II, when possible, was evaluated for these 12 variants. Seven variants were not selected for additional genotyping due to three reasons: high LD with SNPs selected (rs718273, rs10507385), inconsistency of associations with breast cancer when stratified by SBCS study population (rs17063612, rs4771249, rs9508016, rs12429309), or high LD with variants that had inconsistent associations when stratified by SBCS study population (rs3794400). Five SNPs were selected for additional genotyping in Stage 2 including one in VEGFA (rs8330070), three in FLT1 (rs9551471, rs3812867, and rs9554312), and one in KDR (rs10006115). Associations with breast cancer risk for these 5 SNPs are presented in Table 3. Three SNPs showing significant associations in Stage 1 were not replicated in Stage 2 (rs3812867, rs9554312, and rs10006115), while two SNPs showed associations in both genotyping Stages that were generally consistent (rs833070 and rs9551471). Combined analyses of Stage 1 and 2 data revealed an uncorrected significant association of breast cancer risk for VEGFA rs833070 in recessive models, and for FLT1 rs9551471 regardless of genetic model. Permutation tests were employed to evaluate the corrected statistical significance of these associations. The permutation test corrected recessive p-value for VEGFA rs833070 was 0.13, and the permutation test corrected allelic p-value for FLT1 rs9551471 was 0.02.
Table 3. Selected VEGF Family SNPs and Breast Cancer Risk, the Shanghai Breast Cancer Genetics Study.
| Gene, SNP, SNP Information * | N genotyped (cases / controls) |
Genotype Frequencies ** | Breast Cancer Risk, Additive Models 1 | Dominant Models 2 | Recessive Models 3 | ||||
|---|---|---|---|---|---|---|---|---|---|
|
|
|
|
|||||||
| AB OR (95% CI) | BB OR (95% CI) | p-value | AB/BB OR (95% CI) | p-value | BB OR (95% CI) | p-value | |||
| VEGFA rs833070 (C/T), 22.6%, intron 2 | |||||||||
| Genotyping Stage 1 | 2,066 / 2,141 | 58.3 / 37.0 / 4.7 | 0.96 (0.84-1.09) | 1.35 (1.03-1.77) | 0.378 | 1.00 (0.89-1.13) | 0.994 | 1.37 (1.05-1.79) | 0.022 |
| Genotyping Stage 2 | 4,329 / 1,796 | 60.9 / 34.4 / 4.7 | 1.05 (0.93-1.18) | 1.18 (0.91-1.53) | 0.189 | 1.06 (0.95-1.19) | 0.284 | 1.16 (0.90-1.50) | 0.251 |
| Combined | 6,395 / 3,937 | 59.5 / 35.8 / 4.7 | 1.01 (0.92-1.10) | 1.26 (1.05-1.53) | 0.107 | 1.04 (0.95-1.13) | 0.398 | 1.26 (1.05-1.52) | 0.016 |
| permutation corrected recessive p-value = 0.13 | |||||||||
| FLT1 rs9551471 (A/G), 22.7%, intron 6 | |||||||||
| Genotyping Stage 1 | 2,073 / 2,081 | 59.1 / 36.0 / 4.9 | 0.87 (0.76-0.99) | 0.81 (0.61-1.09) | 0.018 | 0.86 (0.76-0.97) | 0.018 | 0.86 (0.64-1.15) | 0.296 |
| Genotyping Stage 2 | 4,301 / 1,786 | 60.5 / 34.2 / 5.3 | 0.96 (0.85-1.08) | 0.75 (0.58-0.98) | 0.081 | 0.93 (0.83-1.05) | 0.231 | 0.76 (0.59-0.99) | 0.039 |
| Combined | 6,374 / 3,867 | 61.3 / 34.2 / 4.5 | 0.92 (0.84-1.00) | 0.78 (0.64-0.95) | 0.005 | 0.90 (0.83-0.98) | 0.016 | 0.80 (0.66-0.97) | 0.025 |
| permutation corrected allelic p-value = 0.02 | |||||||||
| FLT1 rs3812867 (G/A), 9.4%, exon 30 (3′ UTR) | |||||||||
| Genotyping Stage 1 | 2,068 / 2,145 | 82.8 / 16.7 / 0.5 | 0.95 (0.81-1.12) | 2.26 (1.07-4.78) | 0.741 | 0.99 (0.84-1.16) | 0.877 | 2.28 (1.07-4.82) | 0.032 |
| Genotyping Stage 2 | 4,354 / 1,794 | 81.6 / 17.6 / 0.9 | 0.90 (0.78-1.05) | 0.82 (0.45-1.51) | 0.139 | 0.90 (0.78-1.04) | 0.147 | 0.84 (0.46-1.53) | 0.566 |
| Combined | 6,422 / 3,939 | 82.7 / 16.5 / 0.8 | 0.92 (0.83-1.03) | 1.26 (0.78-2.03) | 0.352 | 0.94 (0.84-1.04) | 0.221 | 1.27 (0.79-2.06) | 0.324 |
| FLT1 rs9554312 (A/G), 30.2%, 3′ FR § | |||||||||
| Genotyping Stage 1 | 1,587 / 1,623 | 51.9 / 38.5 / 9.6 | 1.21 (1.04-1.40) | 1.24 (0.98-1.58) | 0.010 | 1.21 (1.06-1.39) | 0.006 | 1.14 (0.91-1.44) | 0.249 |
| Genotyping Stage 2 | 4,344 / 1,798 | 47.2 / 42.8 / 10.1 | 0.95 (0.85-1.07) | 0.91 (0.75-1.11) | 0.272 | 0.94 (0.85-1.06) | 0.311 | 0.93 (0.78-1.12) | 0.469 |
| Combined | 5,931 / 3,421 | 48.6 / 41.6 / 9.9 | 1.04 (0.95-1.14) | 1.03 (0.88-1.19) | 0.476 | 1.04 (0.95-1.13) | 0.377 | 1.01 (0.87- 1.16) | 0.924 |
| KDR rs10006115 (G/T), 9.4%, intron 29 | |||||||||
| Genotyping Stage 1 | 1,917 / 1,924 | 83.3 / 16.0 / 0.8 | 1.23 (1.04-1.46) | 0.98 (0.47-2.03) | 0.028 | 1.22 (1.04-1.44) | 0.017 | 0.94 (0.45-1.95) | 0.870 |
| Genotyping Stage 2 | 4,340 / 1,797 | 80.4 / 19.0 / 0.7 | 0.91 (0.79-1.05) | 1.18 (0.61-2.30) | 0.345 | 0.92 (0.80-1.06) | 0.260 | 1.20 (0.62-2.34) | 0.586 |
| Combined | 6,257 / 3,721 | 81.6 / 17.7 / 0.8 | 1.03 (0.93-1.15) | 1.08 (0.67-1.76) | 0.521 | 1.03 (0.93-1.15) | 0.538 | 1.08 (0.66-1.75) | 0.765 |
SNP information includes alleles (Major or reference allele / minor or test allele) as determined by allele frequency among all genotyped controls, minor allele frequency among all genotyped controls, and region of the gene where SNP is located
Genotype Frequencies for AA (major allele homozygotes) / AB (heterozygotes) / BB (minor allele homozygotes) among controls
Breast Cancer Risk for heterozygotes (AB) and homozygotes (BB), each compared to major allele homozygotes (AA) in models adjusted for age, education, and genotyping stage when appropriate; p-value for trend
Breast Cancer Risk for minor alelle carriers (AB/BB) compared to major allele homozygotes (AA) in models adjusted for age, education, and genotyping stage when appropriate; p-value for dominant association
Breast Cancer Risk for minor allele homozygotes (BB) compared to major allele carriers (AA/AB) in models adjusted for age, education, and genotyping stage when appropriate; p-value for recessive association
3′ FR: 3′ flanking region, downstream of the gene
Estimates and p-values in bold are significant at p≤0.05
The effects associated with VEGFA rs833070 and FLT1 rs9551471 were found to be independent (p-value for interaction = 0.505), and so a combined analysis was conducted by summing the total number of risk alleles/genotypes (Table 4). To accommodate differences in types and directions of associations with breast cancer risk, a risk allele/genotype was defined as a minor allele homozygote for VEGFA rs833070 and increasing copies of the major allele for FLT1 rs9551471. Compared to women without any risk alleles/genotypes (4.3%), those with one (32.5%), two (60.0%), or three (3.2%) risk alleles/genotypes were greater than 30% (OR: 1.33, 95%CI: 1.13-1.57), 40% (OR: 1.45, 95% CI: 1.24-1.70), and 80% (OR: 1.88, 95% CI: 1.42-2.50) more likely to be breast cancer cases, respectively. This association followed a significant linear trend (p-value = 0.0004). Additional adjustment for menopausal status, hormone replacement therapy, and regular physical activity did not substantially alter these results (data not shown). When stratified, no effect measure heterogeneity was found by menopausal status or family history of breast cancer; multiplicative interaction terms were not statistically significant. Additionally, cases were stratified by tumor stage, and ER or PR status. The risk of breast cancer associated with VEGFA rs833070 and FLT1 rs9551471 was generally consistent by disease stage and ER or PR status, although effects seemed to be more pronounced among early stage patients, as well among patients with PR negative tumors (Table 4).
Table 4. FLT1 (VEGFR1) rs9551471 and VEGFA rs833070 and Breast Cancer Risk, the Shanghai Breast Cancer Genetics Study.
| Analysis | Breast Cancer Risk by Risk Alleles / Genotypes of VEGFR1 / FLT1 rs9551471 and VEGFA rs833070 * | p-value for trend | |||
|---|---|---|---|---|---|
|
| |||||
| 0 (254 / 188) | 1 (2,014 / 1,299) | 2 (3,849 / 2,267) | 3 (225 / 102) | ||
| All Women | 1.0 (reference) | 1.33 (1.13-1.57) | 1.45 (1.24-1.70) | 1.88 (1.42-2.50) | 0.0004 |
| Menopausal Status | |||||
| Premenopausal | 1.0 (reference) | 1.21 (0.97-1.52) | 1.41 (1.14-1.75) | 1.99 (1.32-2.99) | 0.0090 |
| Postmenopausal | 1.0 (reference) | 1.45 (1.13-1.86) | 1.47 (1.16-1.87) | 1.87 (1.25-2.78) | 0.0132 |
| p-value for interaction = 0.87 | |||||
| Tumor Stage | |||||
| Early Stage (TNM <3) | 1.0 (reference) | 1.36 (1.14-1.63) | 1.48 (1.25-1.76) | 1.97 (1.46-2.66) | 0.0006 |
| Late Stage (TNM ≥3) | 1.0 (reference) | 1.45 (1.13-1.86) | 1.47 (1.16-1.87) | 1.87 (1.25-2.78) | 0.0132 |
| p-value for case-case comparison = 0.81 | |||||
| ER/PR Status | |||||
| ER positive PR positive | 1.0 (reference) | 1.31 (1.02-1.69) | 1.39 (1.09-1.77) | 1.80 (1.20-2.70) | 0.1411 |
| ER positive PR negative | 1.0 (reference) | 1.14 (0.76-1.72) | 1.40 (0.95-2.06) | 2.06 (1.11-3.83) | 0.0161 |
| ER negative PR positive | 1.0 (reference) | 1.36 (0.97-1.91) | 1.31 (0.95-1.82) | 1.98 (1.16-3.37) | 0.1867 |
| ER negative PR negative | 1.0 (reference) | 1.41 (1.15-1.73) | 1.54 (1.26-1.87) | 2.07 (1.48-2.90) | 0.0002 |
| p-value for case-case comparison = 0.73 | |||||
Number of risk alleles/genotyped defined as: rs9551471 A and rs833070 TT, (N cases / N controls)
Odds ratio and 95% confidence interval for the risk of breast cancer, adjusted for age, education, and genotyping stage
Estimates and p-values in bold are significant at p≤0.05
Discussion
Inherited genetic variation in two VEGF family ligands (VEGFA and VEGFC) and two VEGF family receptors (FLT1 and KDR) was comprehensively characterized among participants of the Shanghai Breast Cancer Genetics Study. Two SNPs were found to have consistent relationships with breast cancer risk. Minor allele homozygotes of VEGFA rs833070 had an increased risk of breast cancer, while the minor allele of FLT1 rs9551471 was associated with decreasing breast cancer risk in a dose-response manner. To our knowledge, neither of these polymorphisms has been previously evaluated in relation to cancer susceptibility.
Previous studies of genetic variation in VEGF gene family members and breast cancer risk have included a limited number of polymorphisms with mixed findings. VEGFA SNPs discussed, including their names and relative locations, are shown in Figure 1. Two VEGFA SNPs (rs1005230 and rs10434) have had null results in a single study (17), while three SNPs (rs1570360, rs2010963, and rs25648), have consistently not been associated with breast cancer risk in several studies (10-14, 17, 18). On the contrary, three SNPs (rs699947, rs833061, and rs3025039) have had inconsistent results. The rs699947 CC genotype was associated with a significantly increased risk of breast cancer (OR: 1.99, 95% CI: 1.06-3.74) in one study (18), but not associated in four other studies (11, 12, 14, 17). Similarly, the rs833061 CC genotype was associated with a significantly increased risk (OR: 2.01, 95% CI: 1.08-3.76) in one study (18), but not associated with breast cancer risk in four other studies (12-14, 17). Three studies have reported protective effects for rs3025039: the T allele was associated with a decreased risk of breast cancer among 1,000 Caucasians (OR: 0.51, 95% CI: 0.38-0.70) (9) and among 457 Caucasian BRCA1 mutation carriers (OR: 0.63, 95% CI: 0.41-0.98) (16), while premenopausal Chinese women with the TT genotype had a reduced risk of breast cancer (OR: 0.45, 95% CI: 0.25-0.79) in our previous report (13). However, additional genotyping among SBCS Stage 2 participants did not confirm such an association (data not shown). This is in agreement with others who initially found a reduced risk associated with the T allele (9), but were unable to replicate findings in a second study population (17). Additionally, one small study found that the T allele occurred more frequently among breast cancer cases (15), while three other studies found no association (11, 12, 18). Previous reports on other VEGF family polymorphisms and breast cancer risk are sparse. Schneider et al. evaluated one FLT1 SNP (-962 C/T) and two KDR SNPs (889 G/A, V2971I, rs2305948, and 1416 A/T, Q472H, rs1870377) and found no association among 520 breast cancer cases and 715 controls (18). Forsti et al. also evaluated these two KDR SNPs, and found no overall association with breast cancer risk (19).
Figure 1. VEGFA SNPs of Interest, The Shanghai Breast Cancer Genetics Study.
Of the VEGF family variants previously evaluated in the breast cancer risk literature, the majority (VEGFA rs699947, rs833061, rs1570360, rs2010963, rs3025039 and KDR rs2305948 and rs1870377) were genotyped in the current study, and not found to be associated with breast cancer susceptibility. Three SNPs previously evaluated in only one study with null findings (17) were not genotyped in the current report (rs1005230, rs25648, and rs10434). Given our high coverage of genetic variation, it is likely that these SNPs were also not related to breast cancer risk in the current analysis. Several VEGFA variants have been reported to be functional. SNPs in the promoter, 5′ UTR, and 3′ UTR have been found to influence expression, although reports for both SNPs (22, 23) and haplotypes (24, 25) have been inconsistent. Associations between VEGFA polymorphisms and plasma or serum levels have also produced mixed findings (9, 14, 17, 26, 27). In the current study, four potentially functional VEGFA SNPs were genotyped (rs699947, rs1570360, rs2010963, and rs3025039); none were found be significantly associated with breast cancer risk. An 18 bp insertion-deletion in the VEGFA promoter has also had inconsistent associations with VEGFA expression (23, 28, 29). This insertion/deletion polymorphism is reported to be in perfect LD with rs699947 (30), which was genotyped in our study and not found to be associated with susceptibility to breast cancer. However, rs699947 is in high LD with rs833070 (D′ = 1.0, r2 = 0.84), so rs833070 may also capture some of the variation due to the insertion-deletion polymorphism.
VEGFA is the primary mitogen of the VEGF family, and angiogenic signaling is normally mediated by interactions with KDR (VEGFR2) (1-3). However, during pathogenic angiogenesis, signaling via FLT1 (VEGFR1) may be more important (1, 8); for example, FLT1 contributes to the development of the ‘premetastatic niche’ whereas KDR does not (36). This is consistent with our finding of SNPs in VEGFA and FLT1, but not KDR, to be significantly associated with altered breast cancer susceptibility. Both polymorphisms found to be associated with breast cancer risk are located in introns; rs833070 is in intron 2 of VEGFA and rs9551471 is in intron 6 of FLT1. In silico analysis was conducted to assess the potential functional importance of these loci. No regulatory features were found surrounding or immediately adjacent to rs9551471 in the Ensembl genome browser (37), while rs833070 was found to be located within a large (1.8 kb) gene associated putative regulatory element that is enriched with CTCF and DNaseI sites. In agreement with this, FASTSNP (38) scored rs9551471 as 0 (intronic with no known function) and rs833070 as 1-2 (possible intronic enhancer). As it is possible that these SNPs are themselves not functional, but are in high LD with the true casual variants, SNPs in very high LD were also evaluated; no additional insights were revealed. Thus, we can conjecture that these intronic SNPs possibly influence gene expression or mRNA splicing, or are in LD with ungenotyped variations that do. Notably, VEGFA rs833070 has recently been found to be associated with two other phenotypes. In a small study of healthy adults, rs833070 minor allele carriers had significantly reduced hippocampus concentration as assessed by high-resolution structural magnetic resonance imaging scans (39), and among 175 Japanese patients with type 1 diabetes, VEGFA rs833070 minor allele homozygotes had a significantly shorter time to nonproliferative diabetic retinopathy (NPDR) progression than major allele carriers (30). These findings highlight the diverse roles of the VEGF family, which is now known to extend far beyond angiogenesis.
Strengths of the current study include our comprehensive SNP coverage, a very large study population, and a multigenic approach to assessing breast cancer risk. To our knowledge, this is the largest and most comprehensive study on VEGF family polymorphisms and breast cancer to date. In summary, our data indicate that common genetic variation in VEGF family genes may contribute to altered breast cancer susceptibility among Chinese women. Specifically, we found that women with the VEGFA rs833070 TT genotype and/or increasing copies of the FLT1 (VEGFR1) rs9551471 A allele, had increasing risks of breast cancer in a dose-response manner. These findings are consistent with recent meta-analyses that reported no significant associations between VEGFA -2578 C/A (rs699947), -460C/T (rs833061), and +936 C/T (rs3025039) variants and breast cancer risk (31-35). To date, no genome wide association studies (GWAS) of breast cancer have detected significant associations between VEGF family variants and breast cancer risk. Given the p-value threshold required by GWAS, and the possibly lower coverage of genetic variation, these results are not inconsistent with our findings. Additional studies are needed to confirm our findings and evaluate genetic variation in other VEGF family members, including VEGF-B, FIGF (VEGF-D), PGF, and FLT4 (VEGFR3), in relation to breast cancer risk.
Supplementary Material
Acknowledgments
The authors are grateful to the study participants and research staff of the Shanghai Breast Cancer Genetics Study. This research was supported in part by US National Institutes of Health grants R01CA64277, R01CA70867, R01CA90899, and R01CA124558 to W.Z., R01 CA118229, R01CA92585 and Department of Defense (DOD) Idea Award BC011118 to X.O.S., and R01CA122756 and DOD Idea Award BC050791 to Q.C. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institute of Health. We would like to thank Ms. Regina Courtney and Dr. Shawn Levy for their contributions to the genotyping, Dr. Ryan Delahanty and Dr. Todd Edwards for their discussions and tutorials on permutation tests, and Ms. Brandy Venuti and Ms. Mary Jo Daly for their assistance in the preparation of this manuscript. Genotyping assays were conducted at the Survey and Biospecimen Shared Resource (TaqMan), Vanderbilt Microarray Shared Resource (Affymetrix Targeted Genotyping System and Genome-Wide SNP Array 6.0), and Proactive Genomics (Sequenom). The Survey and Biospecimen Shared Resource and Vanderbilt Microarray Shared Resource are supported in part by the Vanderbilt-Ingram Cancer Center (P30 CA68485).
References
- 1.Takahashi H, Shibuya M. The vascular endothelial growth factor (VEGF)/VEGF receptor system and its role under physiological and pathological conditions. Clin Sci (Lond) 2005;109:227–41. doi: 10.1042/CS20040370. [DOI] [PubMed] [Google Scholar]
- 2.Roy H, Bhardwaj S, Yla-Herttuala S. Biology of vascular endothelial growth factors. FEBS Lett. 2006;580:2879–87. doi: 10.1016/j.febslet.2006.03.087. [DOI] [PubMed] [Google Scholar]
- 3.Roskoski R., Jr Vascular endothelial growth factor (VEGF) signaling in tumor progression. Crit Rev Oncol Hematol. 2007;62:179–213. doi: 10.1016/j.critrevonc.2007.01.006. [DOI] [PubMed] [Google Scholar]
- 4.Byrne AM, Bouchier-Hayes DJ, Harmey JH. Angiogenic and cell survival functions of vascular endothelial growth factor (VEGF) J Cell Mol Med. 2005;9:777–94. doi: 10.1111/j.1582-4934.2005.tb00379.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kowanetz M, Ferrara N. Vascular endothelial growth factor signaling pathways: therapeutic perspective. Clin Cancer Res. 2006;12:5018–22. doi: 10.1158/1078-0432.CCR-06-1520. [DOI] [PubMed] [Google Scholar]
- 6.Breen EC. VEGF in biological control. J Cell Biochem. 2007;102:1358–67. doi: 10.1002/jcb.21579. [DOI] [PubMed] [Google Scholar]
- 7.Epstein RJ. VEGF signaling inhibitors: more pro-apoptotic than anti-angiogenic. Cancer Metastasis Rev. 2007;26:443–52. doi: 10.1007/s10555-007-9071-1. [DOI] [PubMed] [Google Scholar]
- 8.Shibuya M. Vascular endothelial growth factor receptor-1 (VEGFR-1/Flt-1): a dual regulator for angiogenesis. Angiogenesis. 2006;9:225–30. doi: 10.1007/s10456-006-9055-8. [DOI] [PubMed] [Google Scholar]
- 9.Krippl P, Langsenlehner U, Renner W, et al. A common 936 C/T gene polymorphism of vascular endothelial growth factor is associated with decreased breast cancer risk. Int J Cancer. 2003;106:468–71. doi: 10.1002/ijc.11238. [DOI] [PubMed] [Google Scholar]
- 10.Smith KC, Bateman AC, Fussell HM, Howell WM. Cytokine gene polymorphisms and breast cancer susceptibility and prognosis. Eur J Immunogenet. 2004;31:167–73. doi: 10.1111/j.1365-2370.2004.00462.x. [DOI] [PubMed] [Google Scholar]
- 11.Jin Q, Hemminki K, Enquist K, et al. Vascular endothelial growth factor polymorphisms in relation to breast cancer development and prognosis. Clin Cancer Res. 2005;11:3647–53. doi: 10.1158/1078-0432.CCR-04-1803. [DOI] [PubMed] [Google Scholar]
- 12.Jacobs EJ, Feigelson HS, Bain EB, et al. Polymorphisms in the vascular endothelial growth factor gene and breast cancer in the Cancer Prevention Study II cohort. Breast Cancer Res. 2006;8:R22. doi: 10.1186/bcr1400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kataoka N, Cai Q, Wen W, et al. Population-based case-control study of VEGF gene polymorphisms and breast cancer risk among Chinese women. Cancer Epidemiol Biomarkers Prev. 2006;15:1148–52. doi: 10.1158/1055-9965.EPI-05-0871. [DOI] [PubMed] [Google Scholar]
- 14.Balasubramanian SP, Cox A, Cross SS, Higham SE, Brown NJ, Reed MW. Influence of VEGF-A gene variation and protein levels in breast cancer susceptibility and severity. Int J Cancer. 2007;121:1009–16. doi: 10.1002/ijc.22772. [DOI] [PubMed] [Google Scholar]
- 15.Eroglu A, Ozturk A, Cam R, Akar N. Vascular endothelial growth factor gene 936 C/T polymorphism in breast cancer patients. Med Oncol. 2008;25:54–55. doi: 10.1007/s12032-007-0046-4. [DOI] [PubMed] [Google Scholar]
- 16.Jakubowska A, Gronwald J, Menkiszak J, et al. The VEGF_936_C>T 3′UTR polymorphism reduces BRCA1-associated breast cancer risk in Polish women. Cancer Letters. 2007;262:71–76. doi: 10.1016/j.canlet.2007.11.029. [DOI] [PubMed] [Google Scholar]
- 17.Langsenlehner U, Wolf G, Langsenlehner T, et al. Genetic polymorphisms in the vascular endothelial growth factor gene and breast cancer risk. The Austrian “tumor of breast tissue: incidence, genetics, and environmental risk factors” study. Breast Cancer Res Treat. 2008;109:297–304. doi: 10.1007/s10549-007-9655-z. [DOI] [PubMed] [Google Scholar]
- 18.Schneider BP, Radovich M, Sledge GW, et al. Association of polymorphisms of angiogenesis genes with breast cancer. Breast Cancer Res Treat. 2008;111:157–63. doi: 10.1007/s10549-007-9755-9. [DOI] [PubMed] [Google Scholar]
- 19.Forsti A, Jin Q, Altieri A, et al. Polymorphisms in the KDR and POSTN genes: association with breast cancer susceptibility and prognosis. Breast Cancer Res Treat. 2007;101:83–93. doi: 10.1007/s10549-006-9265-1. [DOI] [PubMed] [Google Scholar]
- 20.Zheng W, Long J, Gao YT, et al. Genome-wide association study identifies a new breast cancer susceptibility locus at 6q25.1. Nat Genet. 2009;41:324–8. doi: 10.1038/ng.318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81:559–75. doi: 10.1086/519795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Shahbazi M, Fryer AA, Pravica V, et al. Vascular endothelial growth factor gene polymorphisms are associated with acute renal allograft rejection. J Am Soc Nephrol. 2002;13:260–4. doi: 10.1681/ASN.V131260. [DOI] [PubMed] [Google Scholar]
- 23.Salvarani C, Boiardi L, Casali B, et al. Vascular endothelial growth factor gene polymorphisms in Behcet's disease. J Rheumatol. 2004;31:1785–9. [PubMed] [Google Scholar]
- 24.Prior SJ, Hagberg JM, Paton CM, et al. DNA sequence variation in the promoter region of the VEGF gene impacts VEGF gene expression and maximal oxygen consumption. Am J Physiol Heart Circ Physiol. 2006;290:H1848–55. doi: 10.1152/ajpheart.01033.2005. [DOI] [PubMed] [Google Scholar]
- 25.Stevens A, Soden J, Brenchley PE, Ralph S, Ray DW. Haplotype analysis of the polymorphic human vascular endothelial growth factor gene promoter. Cancer Res. 2003;63:812–6. [PubMed] [Google Scholar]
- 26.Renner W, Kotschan S, Hoffmann C, Obermayer-Pietsch B, Pilger EA. Common 936 C/T Mutation in the Gene for Vascular Endothelial Growth Factor Is Associated with Vascular Endothelial Growth Factor Plasma Levels. J Vasc Res. 2000;37:443–8. doi: 10.1159/000054076. [DOI] [PubMed] [Google Scholar]
- 27.Awata T, Inoue K, Kurihara S, et al. A common polymorphism in the 5′-untranslated region of the VEGF gene is associated with diabetic retinopathy in type 2 diabetes. Diabetes. 2002;51:1635–9. doi: 10.2337/diabetes.51.5.1635. [DOI] [PubMed] [Google Scholar]
- 28.Brogan IJ, Khan N, Isaac K, Hutchinson JA, Pravica V, Hutchinson IV. Novel polymorphisms in the promoter and 5′ UTR regions of the human vascular endothelial growth factor gene. Human Immunology. 1999;60:1245–9. doi: 10.1016/s0198-8859(99)00132-9. [DOI] [PubMed] [Google Scholar]
- 29.Yang B, Cross DF, Ollerenshaw M, Millward BA, Demaine AG. Polymorphisms of the vascular endothelial growth factor and susceptibility to diabetic microvascular complications in patients with type 1 diabetes mellitus. Journal of Diabetes and its Complications. 2001;17:1–6. doi: 10.1016/s1056-8727(02)00181-2. [DOI] [PubMed] [Google Scholar]
- 30.Nakanishi K, Watanabe C. Single nucleotide polymorphisms of vascular endothelial growth factor gene intron 2 are markers for early progression of diabetic retinopathy in Japanese with type 1 diabetes. Clin Chim Acta. 2009;402:171–5. doi: 10.1016/j.cca.2009.01.004. [DOI] [PubMed] [Google Scholar]
- 31.Cao C, Ying T, Fang JJ, et al. Polymorphism of vascular endothelial growth factor - 2578C/A with cancer risk: evidence from 11263 subjects. Med Oncol. 2010 doi: 10.1007/s12032-010-9613-1. [DOI] [PubMed] [Google Scholar]
- 32.Gu D, Wang M. VEGF 936C>T polymorphism and breast cancer risk: evidence from 5,729 cases and 5,868 controls. Breast Cancer Res Treat. 2010 doi: 10.1007/s10549-010-0991-z. [DOI] [PubMed] [Google Scholar]
- 33.Xu B, Feng NH, Tong N, et al. VEGF -460C>T polymorphism and cancer risk: a meta-analysis. Med Oncol. 2009 doi: 10.1007/s12032-009-9329-2. [DOI] [PubMed] [Google Scholar]
- 34.Xu B, Li JM, Tong N, et al. VEGFA +936C>T polymorphism and cancer risk: a meta-analysis. Cancer Genet Cytogenet. 2010;198:7–14. doi: 10.1016/j.cancergencyto.2009.11.007. [DOI] [PubMed] [Google Scholar]
- 35.Yang DS, Park KH, Woo OH, et al. Association of a vascular endothelial growth factor gene 936 C/T polymorphism with breast cancer risk: a meta-analysis. Breast Cancer Res Treat. 2010 doi: 10.1007/s10549-010-1070-1. [DOI] [PubMed] [Google Scholar]
- 36.Kaplan RN, Riba RD, Zacharoulis S, et al. VEGFR1-positive haematopoietic bone marrow progenitors initiate the pre-metastatic niche. Nature. 2005;438:820–7. doi: 10.1038/nature04186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Flicek P, Aken BL, Ballester B, et al. Ensembl's 10th year. Nucleic Acids Res. 2010;38:D557–62. doi: 10.1093/nar/gkp972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Yuan HY, Chiou JJ, Tseng WH, et al. FASTSNP: an always up-to-date and extendable service for SNP function analysis and prioritization. Nucleic Acids Res. 2006;34:W635–41. doi: 10.1093/nar/gkl236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Blumberg HP, Wang F, Chepenik LG, et al. Influence of vascular endothelial growth factor variation on human hippocampus morphology. Biol Psychiatry. 2008;64:901–3. doi: 10.1016/j.biopsych.2008.07.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.