Abstract
Purpose
The evidence of a relation between folate intake and one-carbon metabolism (OCM) with pancreatic cancer (PanCa) is inconsistent. In this study, the association between genes and single-nucleotide polymorphisms (SNPs) related to OCM and PanCa was assessed.
Methods
Using biochemical knowledge of the OCM pathway, we identified thirty-seven genes and 834 SNPs to examine in association with PanCa. Our study included 1,408 cases and 1,463 controls nested within twelve cohorts (PanScan). The ten SNPs and five genes with lowest p values (<0.02) were followed up in 2,323 cases and 2,340 controls from eight case-control studies (PanC4) that participated in PanScan2. The correlation of SNPs with metabolite levels was assessed for 649 controls from the European Prospective Investigation into Cancer and Nutrition.
Results
When both stages were combined, we observed suggestive associations with PanCa for rs10887710 (MAT1A) (OR 1.13, 95%CI 1.04-1.23), rs1552462 (SYT9) (OR 1.27, 95%CI 1.02-1.59), and rs7074891 (CUBN) (OR 1.91, 95%CI 1.12-3.26). After correcting for multiple comparisons, no significant associations were observed in either the first or second stage. The three suggested SNPs showed no correlations with one-carbon biomarkers.
Conclusions
This is the largest genetic study to date to examine the relation between germline variations in OCM-related genes polymorphisms and the risk of PanCa. Suggestive evidence for an association between polymorphisms and PanCa was observed among the cohort-nested studies, but this did not replicate in the case-control studies. Our results do not strongly support the hypothesis that genes related to OCM play a role in pancreatic carcinogenesis.
Keywords: Pancreatic cancer, One-carbon metabolism, Polymorphisms, Biomarkers, Epidemiology
Introduction
Impaired DNA methylation is known to cause cancer through induction of chromosomal instability [1]. Folate and related nutrients (homocysteine, cysteine, methionine, cobalamin, and vitamin B6) are thought to influence carcinogenesis through the one-carbon metabolism (OCM) pathway, which is involved in DNA repair, nucleotide synthesis, and methylation [2]. In particular, low folate levels may result in decreased S-adenosyl-methionine levels, the primary donor for methylation reactions. It has been hypothesized that hypomethylation may occur as result of altered folate metabolism [3]. One-carbon metabolites (i.e., folate and related nutrients) have been associated with several types of cancer [4–8] and inverse associations with pancreatic cancer (PanCa) have been observed for dietary folate, although study results have been inconsistent [9–16]. In addition, previous studies in association with PanCa for OCM-related SNPs observed a positive association with PanCa for the MTHFR 677TT genotype [17] and for two SNPs on the MTRR gene (rs162049 and rs10380) [18]. Genome-wide association studies (GWAS) of PanCa have not identified associations with genes involved with OCM [19–24].
In the present study, we used a two-stage approach to evaluate whether 31 genes thought to play a role in the OCM pathway (http://wikipathways.org/index.php/Pathway:WP241) were associated with PanCa. Additionally, genes (ALPL, CPS1, CUBN, FUT2, PRICKLE2, and TCN1) that included SNPs previously observed to predict one-carbon metabolite levels in GWAS [25–27] were also included. The first stage included nested case control studies from twelve cohorts that participated in the National Cancer Institute (NCI) Pancreatic Cancer Cohort Consortium genome-wide association study (PanScan1) [19]. The second stage, used for replication, included eight case-control studies with GWAS data that participated in the Pancreatic Cancer Case-Control Consortium (PanC4) and were part of PanScan 2 [22].
Methods
The methods for PanScan1 and PanScan2 are described elsewhere [19,22]. The study population included PanCa cases and control participants from the previously conducted GWAS in the PanScan1 and PanC4 [19,22]. The first stage of our analyses included 1,408 incident cases and 1,463 controls from the twelve cohort studies that participated in PanScan1 [19]. The second stage included 2,323 cases and 2,340 controls from eight case-control studies that participated in PanC4 and were part of PanScan1 and PanScan2 [19,22]. All included studies had available dietary and lifestyle data. Cases were defined as participants diagnosed with primary adenocarcinoma of the exocrine pancreas and controls were matched to cases according to birth year, sex, and self-reported race/ethnicity and were free of PanCa at the time of recruitment [19,22]. Genotyping was performed by the NCI Core Genotyping Facility using the Illumina HumanHap550 and HumanHap550-Duo SNP arrays (PanScan1) and Illumina Human 610-Quad arrays (PanScan2) [19,22]. Each individual study obtained informed consent from study participants and approval from local institutional review board. The Special Studies Institutional Review Board of the NCI approved the pooled PanScan study [19,22].
Details of the collection of data across the cohorts are described elsewhere [19]. A total of 37 genes and 834 SNPs were included in the study. From all studies, data on the participant’s age and gender were obtained. In addition, five effective principal components or eigenvectors determined from the GWAS data were included as quantitative covariates to correct for population structure.
Data on one-carbon metabolite biomarkers (homocysteine, cysteine, methionine, folate, cobalamin, and vitamin B6) measured in plasma were available for 327 controls of the case-control study nested within the EPIC cohort. Biomarker levels were analyzed by MS-based methods in the laboratory of BEVITAL AS (http://www.bevital.no) in Bergen, Norway [28].
In the first stage, the association for each SNP with PanCa was examined within the cohorts of PanScan using unconditional logistic regression, adjusted for age and sex. To correct for population stratification, analyses were also adjusted for cohort and five principal components of population stratification. Allelic additive models were used where the risk allele was the least frequent allele. p values were adjusted for multiple testing using the permutation-based closed step-down minP procedure [29]. Genes were evaluated separately using the adaptive rank truncated product (ARTP) method [30] with 20 truncation points, using 10,000 permutations. This captures potential multiple association signals within a gene, accounting for gene size and LD structure. Gene ARTP p values were adjusted for multiple comparisons using a false discovery rate [31].
The ten SNPs and five genes that showed lowest p values for the association with PanCa in the cohort data were followed up in eight case-control studies from PanC4, using similar analyses. Because fewer comparisons were performed in this second stage, a Bonferroni correction was applied to both SNP and gene p values to adjust for multiple comparisons. Finally, a combined analysis of both stages was performed. Because combined results were mainly driven by cohort results, these were adjusted for multiple comparisons using similar techniques as applied to the first stage.
Adaptive rank truncated product p values also were computed for the total OCM pathway in all stages using the top 5 genes as truncation points. An adjusted p value below 0.05 was considered statistically significant.
Modification of the top ten SNPs by dietary intake of folate (in tertiles) was assessed in the cohorts by testing the statistical significance of the multiplicative interaction terms using likelihood ratio tests. Differences in least-squares means for plasma metabolite levels (homocysteine, cysteine, methionine, folate, cobalamin, and vitamin B6) were compared per genotype for each SNP by a pairwise t test in controls only.
Results
In the first stage of our analyses (cohort studies), rs10887710 (MAT1A) showed the lowest p value of all tested SNPs (allelic OR 1.24, 95%CI 1.08-1.41, p 0.002) (Table 1). However, after correcting for multiple comparisons, none of the SNPs were statistically significant. None of the top ten SNPs for the first stage showed a relation with PanCa in the second stage (case-control studies), with all p values greater than 0.2. When all studies were combined (cohort and case-control), statistically significant associations were found for rs10887710 (MAT1A), rs1552462 (SYT9), and rs7074891 (CUBN). However, after correcting for multiple comparisons, the associations were not statistically significant.
Table 1.
Top ten SNPs from cohort with corresponding odds ratios and P-values for cohort, case-control and combined data
| Markera, Allelesb, Chrc, Locationc, and Gened | Subsete | MAFf | Subjectsg | P valueh | P value (adjusted for multiple testing)i |
OR(95% CI)j |
|---|---|---|---|---|---|---|
| rs10887710 (T,C), 10q23.1 (82019765), MAT1A | Cohort | 0.17|0.20 | 1463|1407 | 0.002 | 0.564 | 1.24 (1.08-1.41) |
| Replication, case-control | 0.19|0.20 | 2340|2323 | 0.341 | 1.000 | 1.06 (0.94-1.18) | |
| Combined | 0.18|0.20 | 3803|3730 | 0.005 | 0.990 | 1.13 (1.04-1.23) | |
| rs1552462 (C,T), 11p15.4 (7213717), SYT9 | Cohort | 0.02|0.03 | 1463|1408 | 0.010 | 1.000 | 1.58 (1.12-2.23) |
| Replication, case-control | 0.02|0.03 | 2338|2320 | 0.600 | 1.000 | 1.08 (0.81-1.44) | |
| Combined | 0.02|0.03 | 3801|3728 | 0.033 | 1.000 | 1.27 (1.02-1.59) | |
| rs3819101 (C,T), 18p11.32 (667240), TYMS | Cohort | 0.29|0.33 | 1462|1407 | 0.011 | 1.000 | 1.16 (1.03-1.30) |
| Replication, case-control | 0.32|0.33 | 2336|2323 | 0.784 | 1.000 | 1.01 (0.92-1.11) | |
| Combined | 0.31|0.33 | 3798|3729 | 0.098 | 1.000 | 1.06 (0.99-1.14) | |
| rs2276433 (C,T), 11p15.1 (20630443), SLC6A5 | Cohort | 0.38|0.41 | 1463|1407 | 0.013 | 1.000 | 1.15 (1.03-1.28) |
| Replication, case-control | 0.42|0.39 | 2340|2323 | 0.986 | 1.000 | 1.00 (0.89-1.12) | |
| Combined | 0.40|0.40 | 3803|3730 | 0.800 | 1.000 | 0.99 (0.93-1.06) | |
| rs11873007 (C,T), 18p11.32 (670380), TYMS | Cohort | 0.30|0.33 | 1359|1302 | 0.013 | 1.000 | 1.16 (1.03-1.30) |
| Replication, case-control | 0.32|0.34 | 2326|2309 | 0.742 | 1.000 | 1.02 (0.93-1.12) | |
| Combined | 0.31|0.34 | 3685|3611 | 0.100 | 1.000 | 1.06 (0.99-1.14) | |
| rs7074891 (T,C), 10p13 (17228480), CUBN | Cohort | 0.00|0.01 | 1463|1408 | 0.014 | 1.000 | 2.36 (1.22-4.55) |
| Replication, case-control | 0.01|0.01 | 2340|2322 | 0.356 | 1.000 | 1.45 (0.59-3.58) | |
| Combined | 0.01|0.01 | 3803|3730 | 0.017 | 1.000 | 1.91 (1.12-3.26) | |
| rs1835898 (A,G), 3p14.1 (64096330), PRICKLE2 | Cohort | 0.45|0.42 | 1462|1408 | 0.015 | 1.000 | 0.88 (0.79-0.98) |
| Replication, case-control | 0.44|0.44 | 2339|2323 | 0.748 | 1.000 | 1.01 (0.93-1.11) | |
| Combined | 0.44|0.43 | 3801|3731 | 0.172 | 1.000 | 0.95 (0.89-1.02) | |
| rs3740873 (G,T), 11p15.1 (20626016), SLC6A5 | Cohort | 0.45|0.49 | 1462|1407 | 0.016 | 1.000 | 1.14 (1.02-1.27) |
| Replication, case-control | 0.49|0.47 | 2340|2323 | 0.201 | 1.000 | 0.95 (0.87-1.03) | |
| Combined | 0.47|0.48 | 3802|3730 | 0.646 | 1.000 | 1.02 (0.95-1.09) | |
| rs3786355 (C,T), 18p11.32 (671962), TYMS | Cohort | 0.30|0.33 | 1463|1408 | 0.016 | 1.000 | 1.15 (1.03-1.29) |
| Replication, case-control | 0.32|0.33 | 2338|2320 | 0.822 | 1.000 | 1.01 (0.92-1.11) | |
| Combined | 0.31|0.33 | 3801|3728 | 0.125 | 1.000 | 1.06 (0.98-1.14) | |
| rs222338 (C,A), 3p14.1 (64202692), PRICKLE2 | Cohort | 0.05|0.04 | 1463|1408 | 0.017 | 1.000 | 0.73 (0.57-0.95) |
| Replication, case-control | 0.05|0.06 | 2334|2320 | 0.751 | 1.000 | 1.03 (0.85-1.26) | |
| Combined | 0.05|0.05 | 3797|3728 | 0.266 | 1.000 | 0.92 (0.79-1.07) |
The SNP-level results from the unconditional logistic regression of the genotypes generated in a total of 1,407 pancreatic cancer cases and 1,463 controls.
The analysis adjusted for age in ten-year categories, sex, study, five principal components of population stratification.
NCBI dbSNP identifier.
Major allele, minor allele.
Chromosome and NCBI Human genome Build 36 location.
Gene neighborhood within 20 kb upstream and 10 kb downstream of SNP.
Subset; Cohort: Cohort studies, Case-control: Case-Control studies, Combined: all studies.
Minor allele frequency in control and case participants.
Controls, cases.
1 d.f. Wald test.
P-values of cohort and combined stage adjusted for multiple testing by the step-down minP procedure. P-values of case-control stage adjusted by Bonferroni.
Estimate assuming additive model
OR, allelic odds ratio; CI, 95% confidence interval.
Two genes (MAT1A, p 0.021 and TYMS, p 0.033) showed statistically significant ARTP P-values in the first stage (Table 2). In the replication stage, none of the genes showed an association with PanCa. When both stages were combined, MTRR (p for cohort stage: 0.19) showed a significant p value (p 0.05). After adjustment for multiple comparisons, no significant associations were found in all stages. Additionally, the pathway analysis showed ARTP-adjusted p values above 0.50 in all stages.
Table 2.
Top five genes from the cohort, with corresponding min P p-values for cohort, case-control, and combined data, and other genes included in the pathway analysis
| Genea and Chrb | Subsetc | P value (ARTP) | P value (adjusted) |
|---|---|---|---|
| MAT1A, 10q23.1 | Cohort | 0.021 | 0.607 |
| Replication, case-control | 0.255 | 1.000 | |
| Combined | 0.275 | 0.849 | |
| TYMS, 18p11.32 | Cohort | 0.033 | 0.607 |
| Replication, case-control | 0.377 | 1.000 | |
| Combined | 0.815 | 1.000 | |
| GNMT, 6p21.1 | Cohort | 0.132 | 0.863 |
| Replication, case-control | 0.863 | 1.000 | |
| Combined | 0.832 | 1.000 | |
| FUT2, 19q13.33 | Cohort | 0.150 | 0.863 |
| Replication, case-control | 0.546 | 1.000 | |
| Combined | 0.431 | 1.000 | |
| MTRR, 5p15.31 | Cohort | 0.188 | 0.863 |
| Replication, case-control | 0.094 | 0.471 | |
| Combined | 0.048 | 0.595 | |
| SLC6A5, 11p15.1 | Cohort | 0.194 | 0.863 |
| MTHFS, 15q25.1 | Cohort | 0.225 | 0.863 |
| MTHFR, 1p36.3 | Cohort | 0.230 | 0.863 |
| FOLH1, 11p11.2 | Cohort | 0.322 | 0.863 |
| SLC6A9, 1p33 | Cohort | 0.333 | 0.863 |
| GCSH,16q23.2 | Cohort | 0.342 | 0.863 |
| TCN2, 22q12.2 | Cohort | 0.357 | 0.863 |
| CBS, 21q22.3 | Cohort | 0.404 | 0.863 |
| PRICKLE2, 3p14.1 | Cohort | 0.405 | 0.863 |
| NBPF3, 1p36.12 | Cohort | 0.407 | 0.863 |
| AMD1, 6q21 | Cohort | 0.430 | 0.863 |
| CTH, 1p31.1 | Cohort | 0.432 | 0.863 |
| MAT2A, 2p11.2 | Cohort | 0.441 | 0.863 |
| GIF, 11q13 | Cohort | 0.500 | 0.863 |
| CPS1, 2q35 | Cohort | 0.504 | 0.863 |
| SHMT1, 17p11.2 | Cohort | 0.536 | 0.863 |
| FOLR2, 11q13.3-q13.5 | Cohort | 0.553 | 0.863 |
| MTHFD1, 14q24 | Cohort | 0.564 | 0.863 |
| SLC1A5, 19q13.3 | Cohort | 0.590 | 0.863 |
| GGH, 8q12.3 | Cohort | 0.623 | 0.863 |
| ALDH1L1, 3q21.3 | Cohort | 0.644 | 0.863 |
| BHMT, 5q13.1-q15 | Cohort | 0.665 | 0.863 |
| FOLR3, 11q13 | Cohort | 0.675 | 0.863 |
| SHMT2, 12q12-q14 | Cohort | 0.683 | 0.863 |
| SYT9, 11p15.4 | Cohort | 0.700 | 0.863 |
| CUBN, 10p12.31 | Cohort | 0.724 | 0.865 |
| ALPL, 1p36.12 | Cohort | 0.766 | 0.886 |
| TCN1, 11q11-q12 | Cohort | 0.826 | 0.901 |
| SLC19A1, 21q22.3 | Cohort | 0.828 | 0.901 |
| MTR, 1q43 | Cohort | 0.879 | 0.929 |
| DHFR, 5q11.2-q13.2 | Cohort | 0.925 | 0.951 |
| GSS, 20q11.2 | Cohort | 0.983 | 0.983 |
The gene-level results using the Adaptive Rank Truncated Product approach, of the 37 genes in a total of 1,407 pancreatic cancer cases and 1,463 controls. For each SNP, the logistic regression analysis used adjustment for age in ten-year categories, sex, study, five principal components of population stratification. Adjustment for multiple comparisons was done using an FDR correction for the cohort and combined stage, and using a Bonferroni correction for the case-control stage.
Gene neighborhood within 20 kb upstream and 10 kb downstream of SNP.
Chromosome and NCBI Human genome Build 36 location.
Subset; Cohort: Cohort studies, Replication, case-control: Case-control studies, Combined: all studies.
There were no significant interactions of the top one-carbon pathway SNPs or genes by dietary folate. Within the subset of EPIC participants with the one-carbon biomarkers, several associations were observed between one-carbon biomarkers and the ten SNPs with the lowest p values in the cohort stage. For three SNPs in TYMS (rs3819101, rs11873007, and rs3786355) that were in high linkage disequilibrium, the TT genotype correlated with higher methionine levels than the CT (28.46 vs. 25.34 μmol/L; p 0.01) or CC genotype (25.45 μmol/L; p 0.01). The AA genotype at rs1835898 (PRICKLE2) was associated with higher folate (17.26 vs. 14.55 nmol/L for AG; p 0.04) and higher pyridoxal phosphate (PLP) concentrations (63.44 vs. 44.45 nmol/L for AG; p 0.02 and 41.35 nmol/L for GG, p 0.02), when compared to the AG or GG genotypes. The AC genotype at rs222338 (PRICKLE2) was associated with higher cobalamin levels than the CC genotype (477.45 vs. 379.26 pmol/L; p 0.03) (data not shown).
Discussion
This is the largest genetic study to date examining the relation between germline variations in OCM-related genes and PanCa risk. In the first stage of this study (cohort studies), evidence for an association between genes (MAT1A and TYMS) and SNPs (most notably rs10887710 [MAT1A]) involved in OCM with PanCa was observed. However, after correcting for multiple comparisons, these associations were not statistically significant, nor did the suggestive evidence replicate in the second stage (case-control studies) of this study. No interaction between dietary folate and any of the ten SNPs with lowest p values in the first stage was observed.
For several of the top ten SNPs, a correlation with one-carbon biomarkers was observed. This indicates that the SNPs possibly directly influence biomarker levels, which could strengthen evidence for a relation between these SNPs and PanCa. However, none of the SNPs replicated in a second stage and correlations with biomarkers could only be calculated from a subset of all included participants. Additionally, no clear association between one-carbon metabolites and risk of PanCa has been observed within EPIC using the same data on metabolites and PanCa [32].
The reason for observing no replication of the suggestive evidence observed in the first stage of the study may have been due to differences in study designs, as cohort studies may be less prone to survival bias. However, considering the large number of SNPs tested, it is most likely that the significant SNPs observed in the cohort studies, which did not replicate in case-control studies, were false-positives in the first stage of the analysis.
In conclusion, we observed suggestive associations between SNPs and genes involved in one-carbon metabolism and risk of PanCa in the first stage of this study (cohort studies), but none of the results replicated in the second stage (case-control studies). Our results do not strongly support the hypothesis that genes related to OCM play a role in pancreatic carcinogenesis.
Supplementary Material
Acknowledgments
This research was supported by the Intramural Research Program of the National Institutes of Health, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Department of Health and Human Services. This study has been funded by the World Cancer Research Fund (Grant no. 2008/51 to PV). For full acknowledgments, please see electronic supplementary material.
Footnotes
Conflict of interest: none
Literature
- 1.Ehrlich M. DNA methylation in cancer: too much, but also too little. Oncogene. 21(35):5400–5413. doi: 10.1038/sj.onc.1205651. [DOI] [PubMed] [Google Scholar]
- 2.Tibbetts AS, Appling DR. Compartmentalization of Mammalian folate-mediated one-carbon metabolism. Annu Rev Nutr. 30:57–81. doi: 10.1146/annurev.nutr.012809.104810. [DOI] [PubMed] [Google Scholar]
- 3.Liu JJ, Ward RL. Folate and one-carbon metabolism and its impact on aberrant DNA methylation in cancer. Adv Genet. 71:79–121. doi: 10.1016/B978-0-12-380864-6.00004-3. [DOI] [PubMed] [Google Scholar]
- 4.Larsson SC, Orsini N, Wolk A. Vitamin B6 and risk of colorectal cancer: a meta-analysis of prospective studies. JAMA. 303(11):1077–1083. doi: 10.1001/jama.2010.263. [DOI] [PubMed] [Google Scholar]
- 5.Kim D-H, Smith-Warner SA, Spiegelman D, et al. Pooled analyses of 13 prospective cohort studies on folate intake and colon cancer. Cancer Causes Control. 21(11):1919–1930. doi: 10.1007/s10552-010-9620-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Collin SM, Metcalfe C, Refsum H, et al. Circulating folate, vitamin B12, homocysteine, vitamin B12 transport proteins, and risk of prostate cancer: a case-control study, systematic review, and meta-analysis. Cancer Epidemiol Biomarkers Prev. 19(6):1632–1642. doi: 10.1158/1055-9965.EPI-10-0180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Larsson SC, Giovannucci E, Wolk A. Folate and risk of breast cancer: a meta-analysis. J Natl Cancer Inst. 99(1):64–76. doi: 10.1093/jnci/djk006. [DOI] [PubMed] [Google Scholar]
- 8.Kim YI. Folate and carcinogenesis: evidence, mechanisms, and implications. J Nutr Biochem. 10(2):66–88. doi: 10.1016/s0955-2863(98)00074-6. [DOI] [PubMed] [Google Scholar]
- 9.Baghurst PA, McMichael AJ, Slavotinek AH, Baghurst KI, Boyle P, Walker AM. A case-control study of diet and cancer of the pancreas. Am J Epidemiol. 134(2):167–179. doi: 10.1093/oxfordjournals.aje.a116069. [DOI] [PubMed] [Google Scholar]
- 10.Bravi F, Polesel J, Bosetti C, et al. Dietary intake of selected micronutrients and the risk of pancreatic cancer: an Italian case-control study. Ann Oncol. 22(1):202–206. doi: 10.1093/annonc/mdq302. [DOI] [PubMed] [Google Scholar]
- 11.Gong Z, Holly EA, Bracci PM. Intake of folate, vitamins B6, B12 and methionine and risk of pancreatic cancer in a large population-based case-control study. Cancer Causes Control. 20(8):1317–1325. doi: 10.1007/s10552-009-9352-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Larsson SC, Håkansson N, Giovannucci E, Wolk A. Folate intake and pancreatic cancer incidence: a prospective study of Swedish women and men. J Natl Cancer Inst. 98(6):407–413. doi: 10.1093/jnci/djj094. [DOI] [PubMed] [Google Scholar]
- 13.Oaks BM, Dodd KW, Meinhold CL, Jiao L, Church TR, Stolzenberg-Solomon RZ. Folate intake, post-folic acid grain fortification, and pancreatic cancer risk in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. Am J Clin Nutr. 91(2):449–455. doi: 10.3945/ajcn.2009.28433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Stolzenberg-Solomon RZ, Pietinen P, Barrett MJ, Taylor PR, Virtamo J, Albanes D. Dietary and other methyl-group availability factors and pancreatic cancer risk in a cohort of male smokers. Am J Epidemiol. 153(7):680–687. doi: 10.1093/aje/153.7.680. [DOI] [PubMed] [Google Scholar]
- 15.Silverman DT, Swanson CA, Gridley G, et al. Dietary and nutritional factors and pancreatic cancer: a case-control study based on direct interviews. J Natl Cancer Inst. 90(22):1710–1719. doi: 10.1093/jnci/90.22.1710. [DOI] [PubMed] [Google Scholar]
- 16.Keszei AP, Verhage BAJ, Heinen MM, Goldbohm RA, van den Brandt PA. Dietary folate and folate vitamers and the risk of pancreatic cancer in the Netherlands cohort study. Cancer Epidemiol Biomarkers Prev. 18(6):1785–1791. doi: 10.1158/1055-9965.EPI-08-1220. [DOI] [PubMed] [Google Scholar]
- 17.Larsson SC, Giovannucci E, Wolk A. Folate intake, MTHFR polymorphisms, and risk of esophageal, gastric, and pancreatic cancer: a meta-analysis. Gastroenterology. 131(4):1271–1283. doi: 10.1053/j.gastro.2006.08.010. [DOI] [PubMed] [Google Scholar]
- 18.Ohnami S, Sato Y, Yoshimura K, et al. His595Tyr polymorphism in the methionine synthase reductase (MTRR) gene is associated with pancreatic cancer risk. Gastroenterology. 135(2):477–488. doi: 10.1053/j.gastro.2008.04.016. [DOI] [PubMed] [Google Scholar]
- 19.Amundadottir L, Kraft P, Stolzenberg-Solomon RZ, et al. Genome-wide association study identifies variants in the ABO locus associated with susceptibility to pancreatic cancer. Nat Genet. 41(9):986–990. doi: 10.1038/ng.429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Diergaarde B, Brand R, Lamb J, et al. Pooling-based genome-wide association study implicates gamma-glutamyltransferase 1 (GGT1) gene in pancreatic carcinogenesis. Pancreatology. 10(2-3):194–200. doi: 10.1159/000236023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Low S-K, Kuchiba A, Zembutsu H, et al. Genome-wide association study of pancreatic cancer in Japanese population. PLoS One. 5(7):e11824. doi: 10.1371/journal.pone.0011824. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Petersen GM, Amundadottir L, Fuchs CS, et al. A genome-wide association study identifies pancreatic cancer susceptibility loci on chromosomes 13q22.1, 1q32.1 and 5p15.33. Nat Genet. 42(3):224–228. doi: 10.1038/ng.522. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Willis JA, Olson SH, Orlow I, et al. A replication study and genome-wide scan of single-nucleotide polymorphisms associated with pancreatic cancer risk and overall survival. Clin Cancer Res. 18(14):3942–3951. doi: 10.1158/1078-0432.CCR-11-2856. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wu C, Miao X, Huang L, et al. Genome-wide association study identifies five loci associated with susceptibility to pancreatic cancer in Chinese populations. Nat Genet. 44(1):62–66. doi: 10.1038/ng.1020. [DOI] [PubMed] [Google Scholar]
- 25.Tanaka T, Scheet P, Giusti B, et al. Genome-wide association study of vitamin B6, vitamin B12, folate, and homocysteine blood concentrations. Am J Hum Genet. 84(4):477–482. doi: 10.1016/j.ajhg.2009.02.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Lange LA, Croteau-Chonka DC, Marvelle AF, et al. Genome-wide association study of homocysteine levels in Filipinos provides evidence for CPS1 in women and a stronger MTHFR effect in young adults. Hum Mol Genet. 19(10):2050–2058. doi: 10.1093/hmg/ddq062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hazra A, Kraft P, Lazarus R, et al. Genome-wide significant predictors of metabolites in the one-carbon metabolism pathway. Hum Mol Genet. 18(23):4677–4687. doi: 10.1093/hmg/ddp428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Midttun Ø, Hustad S, Ueland PM. Quantitative profiling of biomarkers related to B-vitamin status, tryptophan metabolism and inflammation in human plasma by liquid 15 chromatography/tandem mass spectrometry. Rapid Commun Mass Spectrom. 23(9):1371–1379. doi: 10.1002/rcm.4013. [DOI] [PubMed] [Google Scholar]
- 29.Westfall PH, Young SS. Resampling-based multiple testing: examples and methods for p-value adjustment. Wiley, New York, NY: 1993. [Google Scholar]
- 30.Yu K, Li Q, Bergen AW, et al. Pathway analysis by adaptive combination of P-values. Genet Epidemiol. 33(8):700–709. doi: 10.1002/gepi.20422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol. 57(1):289–300. [Google Scholar]
- 32.Chuang S-C, Stolzenberg-Solomon R, Ueland PM, et al. A U-shaped relationship between plasma folate and pancreatic cancer risk in the European Prospective Investigation into Cancer and Nutrition. Eur J Cancer. 47(12):1808–1816. doi: 10.1016/j.ejca.2011.02.007. [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.