ABSTRACT
Background
Epidemiological studies have suggested positive associations for iron and red meat intake with risk of pancreatic ductal adenocarcinoma (PDAC). Inherited pathogenic variants in genes involved in the hepcidin-regulating iron metabolism pathway are known to cause iron overload and hemochromatosis.
Objectives
The objective of this study was to determine whether common genetic variation in the hepcidin-regulating iron metabolism pathway is associated with PDAC.
Methods
We conducted a pathway analysis of the hepcidin-regulating genes using single nucleotide polymorphism (SNP) summary statistics generated from 4 genome-wide association studies in 2 large consortium studies using the summary data-based adaptive rank truncated product method. Our population consisted of 9253 PDAC cases and 12,525 controls of European descent. Our analysis included 11 hepcidin-regulating genes [bone morphogenetic protein 2 (BMP2), bone morphogenetic protein 6 (BMP6), ferritin heavy chain 1 (FTH1), ferritin light chain (FTL), hepcidin (HAMP), homeostatic iron regulator (HFE), hemojuvelin (HJV), nuclear factor erythroid 2-related factor 2 (NRF2), ferroportin 1 (SLC40A1), transferrin receptor 1 (TFR1), and transferrin receptor 2 (TFR2)] and their surrounding genomic regions (±20 kb) for a total of 412 SNPs.
Results
The hepcidin-regulating gene pathway was significantly associated with PDAC (P = 0.002), with the HJV, TFR2, TFR1, BMP6, and HAMP genes contributing the most to the association.
Conclusions
Our results support that genetic susceptibility related to the hepcidin-regulating gene pathway is associated with PDAC risk and suggest a potential role of iron metabolism in pancreatic carcinogenesis. Further studies are needed to evaluate effect modification by intake of iron-rich foods on this association.
Keywords: hepcidin, iron metabolism pathway, pancreatic cancer, genetic susceptibility, epidemiology
Introduction
In the United States, pancreatic cancer is the third leading cause of cancer mortality, and its incidence is increasing (1). Pancreatic ductal adenocarcinoma (PDAC) is the most common type of pancreatic cancer (2). Known risk factors for PDAC include cigarette smoking, excess weight, type 2 diabetes, and heavy alcohol consumption. Genetic susceptibility also plays a role, with an estimated heritability of up to 21% (3).
Epidemiological studies have shown associations between greater consumption of red meat (4–7) and heme iron (8) and increased PDAC risk. Higher serum iron has also been found to be associated with PDAC, although not consistently (9–11). Experimental studies of iron overload support the hypothesis that iron accumulates in pancreatic islets, resulting in reduced insulin secretion, increased pancreatic β-cell death, and decreased pancreatic β-cell function (12, 13) that contributes to diabetes, a known risk factor for PDAC. Free iron catalyzes the generation of reactive oxygen species (ROS), potentially promoting oxidative stress and inflammation (12).
Hepcidin, a 25–amino acid peptide hormone, maintains iron homeostasis and tightly regulates circulating iron by binding to and degrading iron's receptor protein, ferroportin (14). This process inhibits iron absorption in the duodenum, release of recycled iron from macrophages, and release of stored iron from hepatocytes (15). Although hepcidin is primarily synthesized and secreted by the liver, it can also be produced by macrophages, pancreatic islets, and adipose tissue (15). Hepcidin is produced when iron is abundant and in response to inflammation as part of the acute phase response and erythropoiesis (16). The genes involved in hepcidin regulation are highly conserved in vertebrates (17). Hereditary hemochromatosis is caused by mutations in 5 genes, namely hepcidin (HAMP), homeostatic iron regulator (HFE), hemojuvelin (HJV; also known as HFE2), solute carrier family 40 (iron-regulated transporter) member 1 or ferroportin 1 (SLC40A1), and transferrin receptor 2 (TFR2). Mutations in these genes result in insufficient production of hepcidin and promote excess iron absorption from the diet and accumulation of iron in tissues and organs, most notably the liver, pancreas, and heart (18). Hemochromatosis is associated with pancreatogenic diabetes (19) and hepatocellular carcinoma (18).
The goal of the present analysis was to determine whether the hepcidin-regulating iron metabolism pathway as characterized by common variants in hepcidin-regulating genes is associated with risk of PDAC. For this analysis, we focus on genes involved in iron sensing and regulation of dietary iron absorption. Given the role of hepcidin-regulating genes in hemochromatosis and pancreatogenic diabetes, we hypothesize that the hepcidin-regulating gene pathway is associated with PDAC risk.
Methods
The Pancreatic Cancer Cohort Consortium and the Pancreatic Cancer Case–Control Consortium
Our analysis included 9,253 primary PDAC cases (ICD-O-3 code C250–C259) and 12,525 controls that were part of 4 genome-wide association studies (GWAS) conducted in the Pancreatic Cancer Cohort (PanScan) I–III Consortium and the Pancreatic Cancer Case–Control (PanC4) Consortium (20–24) (Supplemental Figure 1). Details of the studies have been previously described (20–23). We only included participants of European ancestry based on the genomic data (20–23) to avoid confounding by population stratification. Pancreatic cancer cases with non-PDAC subtypes (histology types 8150, 8151, 8153, 8155, and 8240) were excluded because their etiologies are thought to be different. The 3 PanScan GWAS included participants from 16 cohorts from the National Cancer Institute (NCI) Cohort Consortium, 9 case–control studies, and 1 case series (Gastrointestinal Cancer Clinic of Dana–Farber Cancer Institute) (22). PanC4 included 9 case–control studies (23). Within the individual studies, controls for PanScan I and II and PanC4 were matched to cases on age, sex, race, area of residence (case–control studies) and/or smoking (Health Professionals Follow-Up Study, Physicians’ Health Study, Nurses’ Health Study, and Women's Health Study only), and incidence density sampled within each respective cohort study (20–23). PanScan III used previously genotyped controls, mostly from cohort studies. Each participating study obtained written informed consent from their participants and approval from their local Institutional Review Board. The NCI's Special Studies Institutional Review Board approved the consortia study.
The genotyping methods, quality control, and imputation for the originating GWAS have been previously described (20–24). The genotyping for PanScan studies was performed at the NCI's Division of Cancer Epidemiology and Genetics’ Cancer Genomics Research Laboratory using Illumina HumanHap series arrays [Illumina HumanHap550 Infinium II array (20) and Human 610-Quad array (21) for PanScan I and II, respectively; Illumina Omni series arrays (OmniExpress, Omni1M, Omni2.5M, and Omni5M) for PanScan III (22)]. The PanC4 study was genotyped on the Illumina HumanOmniExpressExome-8v1 array at the Johns Hopkins Center for Inherited Disease Research (23). Genotype imputation across the 4 study phases was based on the 1000 Genomes Project (phase 3, v1) reference data set using IMPUTE2 (http://mathgen.stats.ox.ac.uk/impute/impute_v2.html) (24, 25). For quality control, single nucleotide polymorphisms (SNPs) with low-quality imputation score (IMPUTE2 INFO score < 0.3) were excluded (24).
Selection of hepcidin regulatory genes and pathway analysis for pancreatic cancer
Regulation of systemic iron occurs through the hepcidin–ferroportin axis (26–28). We selected genes involved in hepcidin regulation through iron-sensing signals (28). These include bone morphogenetic protein 2 (BMP2), bone morphogenetic protein 6 (BMP6), HAMP, HFE, HJV (also known as HFE2), nuclear factor erythroid 2-related factor 2 (NRF2), transferrin receptor 1 (TFR1), and TFR2. We also included 3 genes in the hepcidin–ferroportin complex, namely SLC40A1, ferritin heavy chain 1 (FTH1), and ferritin light chain (FTL). For the gene-based analyses, we mapped SNPs within a genomic region 20 kb upstream and 20 kb downstream of each of the 11 genes. In total, our analysis included 412 SNPs. We excluded SNPs with minor allele frequency <5% and applied linkage disequilibrium (LD) filtering to highly correlated SNP pairs (r2 > 0.81).
Due to the large overlap of variants on the arrays (Illumina HumanHap550 Infinium II, Human 610-Quad) used for PanScan I and II, these studies were combined and jointly analyzed, whereas PanScan III (OmniExpress, Omni1M, Omni2.5M, and Omni5M) and PanC4 (Illumina HumanOmniExpressExome-8v1) were each analyzed separately. The SNP association analysis was conducted using logistic regression with the SNPs coded as the number of effect alleles in the observed genotype or the expected number of effect alleles for imputed genotype [Prob(SNP = effect allele | reference allele) + 2Prob(SNP = effect allele | effect allele)], adjusted for age (10-y categories: (≤50, 51–60, 61–70, 71–80, and ≥81 y)], sex, top eigenvectors for each study phase, study (PanScan), and geographic region (PanScan). We conducted a meta-analysis combining SNP-level summary statistics from the 4 GWAS using an inverse-variance fixed-effects model.
We then performed a gene- and pathway-based analysis using the summary data-based adaptive rank truncated product (sARTP) method, which combines the SNP-level associations across SNPs in a gene or a pathway (29). Signals from ≤5 of the most PDAC-associated SNPs in a gene region were accumulated. The sARTP adjusts for the number of SNPs in a gene and the number of genes in a pathway through a resampling procedure that controls for false positives. The P values of gene- and pathway-level associations were estimated from the resampled null distribution generated from 1 million permutations. We also performed analyses by subgroup, including sex, age at diagnosis with PDAC (<60 vs. ≥ 60 y), and BMI (in kg/m2; <25, 25 to <30, ≥30, and missing). The statistical analyses were performed using the R programing language (version 3.6.3; R Foundation for Statistical Computing). We considered the pathway-level P value < 0.05 to be statistically significant. All tests were 2-sided.
Expression quantitative trait loci analysis and functional annotation
We conducted an exploratory analysis of expression quantitative trait loci (eQTL) data to assess the cis effect of sARTP-selected SNPs in pancreas (n = 305) and other tissues using data from the NIH Genotype-Tissue Expression project (GTEx) v8 (https://www.gtexportal.org/home) and reported expression significant at an false discover rate <0.05 (30). We also examined the regulatory potential of these SNPs (and SNPs in LD) using data and information from HaploReg (31) and RegulomeDB v1.1 (32). If an SNP in our data set was not genotyped or imputed in HaploReg or RegulomeDB v1.1, we selected an alternative SNP in high LD (r2 ≥ 0.90) using LDlink (https://ldlink.nci.nih.gov) (33).
Exploratory analysis with iron and hematologic traits
We also conducted exploratory analyses examining the association between the sARTP-selected SNPs with biomarkers of iron status (serum iron, transferrin, transferrin saturation, and ferritin) and hematologic traits including hemoglobin, hematocrit, RBC count, and RBC distribution width (RDW) using summary statistics from 2 published GWAS (34, 35). We explored the association between the sARTP-selected SNPs and hematologic traits because the majority of iron in the body is contained in the hemoglobin. The iron status biomarkers GWAS consisted of 23,986 individuals of European ancestry (34). The hematologic traits GWAS was from the UK Biobank cohort Gene ATLAS of 452,264 British individuals of European ancestry (35). For SNPs in our data that were not genotyped or imputed in the other 2 data sets (i.e., genotyped on different platforms), we selected an alternative index SNP in high LD (r2 ≥ 0.90) using LDlink (33). We then standardized the β coefficients to create a heatmap for comparing the SNP associations across the iron status biomarkers and hematologic traits. To standardized β values in both data sets, we calculated z scores, dividing the β by the SE. Because we did not have the SE for the UK Biobank GWAS data on blood traits but had the P values, the z score was obtained using the inverse normal distribution function applied to each P value divided by 2. We also conducted exploratory analyses of the hepcidin regulatory gene pathway in relation to serum iron, transferrin saturation, ferritin, and transferrin, respectively, using the SNP-level summary statistics from the Genetics of Iron Status Consortium (29, 34) and sARTP, applying similar methods as described previously. In total, these analyses included up to 180 SNPs.
Results
Baseline characteristics of the PanScan and PanC4 studies are shown in Table 1. Most PDAC cases were diagnosed at age >60 y. The sex and age distributions of the cases compared to controls were similar within PanScan I and II and PanC4. Compared with the cases, the PanScan III controls had a higher proportion of men and participants <70 y old when selected because PanScan III used previously genotyped controls. Overall, cases were more often men (54.1%), with PanC4 having the highest proportion (57.5%) and the PanScan III having the lowest proportion (49.4%) of men.
TABLE 1.
Descriptive characteristics of participants by study phase from the PanScan and PanC4 studies1
| Study phase | Total | ||||
|---|---|---|---|---|---|
| Characteristic | PanScan I case/control 1746/1812 | PanScan II case/control 1768/1841 | PanScan III case/control 1576/5080 | PanC4 case/control 4163/3792 | Combined case/control 9253/12,525 |
| Case diagnosis age >60 y, n (%) | 1375 (78.8) | 1234 (69.8) | 1192 (75.6) | 2868 (68.9) | 6669 (72.1) |
| Age,2n (%) | |||||
| <50 y | 136 (7.8)/80 (4.4) | 132 (7.5)/150 (8.1) | 176 (11.2)/522 (10.3) | 390 (9.4)/419 (11.0) | 834 (9.0)/1171 (9.3) |
| 50–59 y | 235 (13.5)/264 (14.6) | 402 (22.7)/394 (21.4) | 208 (13.2)/1326 (26.1) | 905 (21.7)/950 (25.1) | 1750 (18.9)/2934 (23.4) |
| 60–69 y | 617 (35.3)/690 (38.1) | 619 (35.0)/605 (32.9) | 433 (27.5)/2474 (48.7) | 1474 (35.4)/1266 (33.4) | 3143 (34.0)/5035 (40.2) |
| ≥70 y | 758 (43.4)/778 (42.9) | 615 (34.8)/692 (37.6) | 759 (48.2)/758 (14.9) | 1394 (33.5)/1157 (30.5) | 3526 (38.1)/3385 (27.0) |
| Sex, n (%) | |||||
| Men | 892 (51.1)/925 (51.0) | 945 (53.5)/965 (52.4) | 778 (49.4)/3795 (74.7) | 2395 (57.5)/2106 (55.5) | 5010 (54.1)/7791 (62.2) |
| Women | 854 (48.9)/887 (49.0) | 823 (46.5)/876 (47.6) | 798 (50.6)/1285 (25.3) | 1768 (42.5)/1686 (44.5) | 4243 (45.9)/4734 (37.8) |
PanC4, Pancreatic Cancer Case–Control Consortium; PanScan, Pancreatic Cancer Cohort Consortium.
Age at pancreatic cancer diagnosis or age when selected to be a control. PanScan III used previously genotyped controls.
The hepcidin-regulating iron metabolism pathway was significantly associated with PDAC (P = 0.0028) (Figure 1). The genes (and corresponding SNPs; Table 2) selected by sARTP that contributed the most to the overall association were HJV (rs6424377, rs10910813, rs2027387, rs10910810, rs10910809), TFR2 (rs62482223, rs56328569), TFR1 (rs4927870, rs41297523, rs13093426, rs41299394, rs12487702), BMP6 (rs61668994), and HAMP (rs10419959, rs12981457, rs10421768, rs10424619, rs2284147). In subgroup analyses (Supplemental Table 1), the pathway association tended to be stronger for women (pathway P value = 0.005; HJV), participants aged ≥60 y (pathway P value = 0.007; HJV, BMP6, TFR1, FTL), and participants who were obese (BMI >30; pathway P value = 0.01; TFR2, HJV, FTH1, TFR1), whereas their respective other strata showed no significant pathway associations (pathway P value > 0.05). The pathway was not significant in analyses stratified by median total meat (531 cases, 3,747 controls) or dietary iron intake (402 cases, 3,412 controls) (Supplemental Table 1).
FIGURE 1.
Genes in the hepcidin-regulating iron metabolism pathway associated with PDAC. The red circles are genes selected by sARTP as contributing the most to the pathway association, and black circles are genes not selected by sARTP. HJV, TFR2, TFR1, BMP6, and HAMP contributed to the pathway. The analysis used meta-analysis summary statistic SNP data from 4 genome-wide association studies (9,253 PDAC cases and 12,525 controls), includes 11 genes and 412 SNPs and is adjusted for age, sex, and the top eigenvectors for the PanScan studies and age, sex, and the top eigenvectors for the PanC4 studies. All statistical tests are 2-sided. BMP2, bone morphogenetic protein 2; BMP6, bone morphogenetic protein 6; FTH1, ferritin heavy chain 1; HAMP, hepcidin; HFE, homeostatic iron regulator; HJV, hemojuvelin; NRF2, nuclear factor erythroid 2-related factor 2; PanC4, Pancreatic Cancer Case–Control Consortium; PanScan, Pancreatic Cancer Cohort Consortium; PDAC, pancreatic ductal adenocarcinoma; sARTP, summary data-based adaptive rank truncated product;SLC40A1, ferroportin 1; SNP, single nucleotide polymorphism; TFR1, transferrin receptor 1; TFR2, transferrin receptor 2.
TABLE 2.
Top SNPs (P values < 0.05) for most significant hepcidin-regulating genes in association with PDAC in the PanScan and PanC4 studies1
| Gene in pathway | Location | Selected SNPs | Chr | Position2 | r 2 3 | EA/RA4 | EAF | Allelic OR (95% CI)5 | SNP P value6 |
|---|---|---|---|---|---|---|---|---|---|
| HJV (HFE2) | 1q21.1 | rs6424377 | 1 | 145428400 | Ref | A/G | 0.41 | 0.93 (0.89, 0.97) | 6.14 × 10–4 |
| rs10910813 | 1 | 145426341 | 0.82 | C/T | 0.36 | 0.94 (0.90, 0.98) | 2.82 × 10–3 | ||
| rs2027387 | 1 | 145402014 | 0.57 | G/C | 0.47 | 1.06 (1.02, 1.11) | 3.24 × 10–3 | ||
| rs10910810 | 1 | 145399229 | 0.58 | T/C | 0.47 | 1.06 (1.02, 1.11) | 3.48 × 10–3 | ||
| rs10910809 | 1 | 145395618 | 0.70 | G/A | 0.37 | 0.94 (0.91, 0.99) | 9.35 × 10–3 | ||
| TFR2 | 7q22.1 | rs62482223 | 7 | 100198878 | Ref | A/C | 0.22 | 0.92 (0.88, 0.97) | 1.27 × 10–3 |
| rs56328569 | 7 | 100202219 | 0.97 | G/C | 0.23 | 0.92 (0.88, 0.97) | 2.00 × 10–3 | ||
| TFR1 | 3q29 | rs4927870 | 3 | 195820576 | Ref | A/C | 0.12 | 1.11 (1.04, 1.18) | 1.51 × 10–3 |
| rs41297523 | 3 | 195784648 | 0.64 | C/T | 0.08 | 1.13 (1.05, 1.22) | 1.95 × 10–3 | ||
| rs13093426 | 3 | 195828806 | 0.66 | T/C | 0.08 | 1.13 (1.05, 1.22) | 2.02 × 10–3 | ||
| rs41299394 | 3 | 195775824 | 0.62 | T/C | 0.07 | 1.13 (1.05, 1.23) | 2.18 × 10–3 | ||
| rs12487702 | 3 | 195775991 | 0.62 | A/G | 0.08 | 1.13 (1.04, 1.22) | 2.51 × 10–3 | ||
| BMP6 | 6p24.3 | rs61668994 | 6 | 7733746 | T/C | 0.14 | 1.11 (1.05, 1.19) | 4.95 × 10–4 | |
| HAMP | 19q13.12 | rs10419959 | 19 | 35764705 | Ref | A/G | 0.21 | 1.07 (1.02, 1.13) | 5.53 × 10–3 |
| rs12981457 | 19 | 35766595 | 0.68 | T/C | 0.17 | 1.07 (1.01, 1.13) | 1.77 × 10–2 | ||
| rs10421768 | 19 | 35772899 | 0.89 | G/A | 0.23 | 1.06 (1.01, 1.11) | 1.87 × 10–2 | ||
| rs10424619 | 19 | 35768237 | 0.93 | T/A | 0.22 | 1.06 (1.01, 1.11) | 2.30 × 10–2 | ||
| rs2284147 | 19 | 35765041 | 0.29 | A/G | 0.48 | 1.05 (1.00, 1.09) | 3.34 × 10–2 |
Top SNPs (up to 5 for each gene) from each significant gene in the sARTP gene pathway analysis for hepcidin-regulating genes-PDAC association derived from 9,253 PDAC cases and 12,525 controls. A, adenine; BMP6, bone morphogenetic protein 6; C, cytosine; Chr, chromosome; EA, effect allele; EAF, effect allele frequency; G, guanine; HAMP, hepcidin; HFE, homeostatic iron regulator; HJV, hemojuvelin; PanC4, Pancreatic Cancer Case–Control Consortium; PanScan, Pancreatic Cancer Cohort Consortium; PDAC, pancreatic ductal adenocarcinoma; RA, referent allele; sARTP, summary data-based adaptive rank truncated product; SNP, single nucleotide polymorphism; T, thymine; TFR1, transferrin receptor 1; TFR2, transferrin receptor 2.
Base pair coordinate hg19.
Linkage disequilibrium (r2) values are indicated between SNPs selected by sARTP and are derived from LDlink (https://ldlink.nci.nih.gov/) in European population data.
Effect allele is defined as the minor allele.
OR from unconditional logistic regression adjusted for study, geographical region, age, sex, and to principal components of population substructure in each of the study phases.
Most significant SNP P value based on permutation.
In the exploratory analyses, the hepcidin-regulating iron metabolism pathway was significantly associated with all the iron status biomarkers (Supplemental Table 2; pathway P values < 1.50 × 10–7). All of the PDAC SNPs selected by sARTP for each gene were associated with either biomarkers of iron status or hematologic traits (Supplemental Figure 2, Supplemental Tables 3 and 4). The HJV variants were only associated with hematologic traits (hemoglobin, hematocrit, RBC, and RDW; P values < 0.05 and >5.0 × 10–6). The TFR2 SNPs were positively associated with all the hematologic traits (P values < 5.0 × 10–10) and inversely associated with serum iron and transferrin saturation (P values < 0.05). The TFR1 SNPs were associated inversely with transferrin saturation and positively with ferritin or transferrin (P values < 0.05 and >5.0 × 10–4). The TFR1 SNPs were also associated with RBC count and RDW (P values < 0.05 and >5.0 × 10–11). The sARTP-selected BMP6 SNP was associated with RDW (P value < 0.005 and ≥5.0 × 10–5). The HAMP SNPs were very significantly associated with RBC and RDW (P values < 0.001 and >5.0 × 10–16) and less strongly associated with hemoglobin and hematocrit (P values < 0.05 and ≥0.005).
We examined eQTL results from GTEx (false discovery rate < 0.05) for the most significant sARTP-selected SNPs in pancreas tissue (Table 3) and other tissues (Supplemental Table 5), as well as exploratory functional annotation (Supplemental Table 6). Alleles in 2 correlated SNPs selected for HJV (r2 = 0.70) rs6424377-A compared with the G allele and rs10910809-G compared with the A allele were associated with increased Neuroblastoma breakpoint family member 10 (NBPF10; P values < 3.0 × 10–8) expression in normal pancreas tissue, as well as whole blood, esophagus–mucosa, esophageal–muscularis, esophagus–gastroesophageal junction, stomach, colon–sigmoid, colon–traverse, artery–aorta, artery–tibial, or cell-cultured fibroblasts (P values < 2.0 × 10–6). The T compared to the C allele of rs10910810 selected for HJV was associated with increased expression of long intergenic nonprotein coding RNA 1719 (LINC01719; P value = 5.00 × 10–5) expression in the normal pancreas tissue and decreased NBPF10 expression in the blood (P value = 1.6 × 10–8). Alleles in 4 correlated SNPs selected for HAMP, rs10419959-A, rs12981457-T, rs10421768-G, and rs10424619-T were associated with lower upstream stimulatory factor 2 (USF2) expression in normal pancreas tissue (P values < 1.0 × 10–5) and esophagus–muscularis (P values < 5.0 × 10–6).
TABLE 3.
eQTL for hepcidin iron-regulating gene pathway SNPs in normal pancreas from the GTEx project1
| GTEx pancreas (n = 305) | ||||||
|---|---|---|---|---|---|---|
| Pathway gene | Location | r 2 2 | SNP-EA | eQTL gene3 | P value4 | β (SE)5 |
| HJV | 1q21.1 | Ref | rs6424377-A | NBPF10 | 2.2 × 10–8 | 0.38 (0.06) |
| LINC01719 | 6.7 × 10–6 | 0.34 (0.07) | ||||
| 1q21.1 | 0.58 | rs10910810-T | LINC01719 | 5.0 × 10–5 | 0.31 (0.07) | |
| 1q21.1 | 0.70 | rs10910809-G | NBPF10 | 2.9 × 10–8 | 0.34 (0.07) | |
| HAMP | 19q13.12 | Ref | rs10419959-A | USF2 | 1.0 × 10–5 | -0.21 (0.05) |
| 19q13.12 | 0.68 | rs12981457-T | USF2 | 1.0 × 10–6 | -0.26 (0.05) | |
| 19q13.12 | 0.89 | rs10421768-G | USF2 | 8.3 × 10–6 | -0.20 (0.04) | |
| 19q13.12 | 0.93 | rs10424619-T | USF2 | 1.1 × 10–5 | -0.21 (0.05) | |
Significant SNPs associated with pancreatic ductal adenocarcinoma in the PanScan and PanC4 studies with expression in pancreatic tissue. EA, effect allele; eQTL, expression quantitative trait loci; GTEx, Genotype-Tissue Expression;HAMP, hepcidin; HJV, hemojuvelin; PanC4, Pancreatic Cancer Case–Control Consortium; PanScan, Pancreatic Cancer Cohort Consortium; Ref, reference; SNP, single nucleotide polymorphism.
Linkage disequilibrium (r2) values are derived from LDlink (https://ldlink.nci.nih.gov/) in European population data.
eQTL gene is defined as the gene in close proximity to the SNP.
P value for the eQTL in pancreas has a false discovery rate <0.05.
Effect sizes (β) and SE are computed as the effect of the effect allele relative to the reference allele.
Discussion
We observed a significant association between the combined effects of common variants in the hepcidin-regulating iron metabolism gene pathway and PDAC. The signals contributing the most to the association were from the HJV, TFR2, TFR1, BMP6, and HAMP genes. Hepcidin is encoded by the HAMP gene and is the primary regulator of iron. When iron status is high, it triggers lysosomal degradation of ferroportin encoded by SLC40A1 in the basolateral membrane of enterocytes and plasma membrane of macrophages (27). HAMP expression is regulated by the liver, sensing intracellular and extracellular iron (36). Hepatocyte TFR1, TFR2, and HFE sense extracellular iron, specifically high circulating concentrations of transferrin-bound iron or transferrin saturation, and signal increased HAMP transcription and hepcidin expression (36). Complementary to this iron-sensing mechanism, increased cellular iron stores also induce the production of BMP6, which acts in a paracrine manner to bind to heterodimeric BMP receptors and its coreceptor HJV activating the SMAD pathway, which stimulates HAMP transcription and hepcidin expression (27). Chronically low production of hepcidin can lead to increased blood iron concentrations, iron overload, and iron accumulation in the pancreas (27). Pathogenic variants in HFE [C282Y and p.His63Asp (H63D)] cause type 1 hemochromatosis, the most prevalent form of hemochromatosis. Although the HFE–H63D polymorphism (but not HFE–C282Y) was significantly associated with increased PDAC risk in an Asian population (37, 38), the HFE gene did not contribute to our pathway association. Among the 5 genes most contributing to our association, 4 (HJV, HAMP, TFR2, BMP6) are known to play a role in non-HFE hemochromatosis. Mutations in HJV and HAMP contribute to juvenile or type 2 hemochromatosis, and mutations in TFR2 contribute to young adult type 3 hemochromatosis. BMP6 mutations may contribute to late-onset moderate iron overload and hereditary hemochromatosis (18).
Experimental and PDAC patient survival studies of hepcidin support our findings and may offer insight into biologic mechanisms. In a study of pancreatic cancer tissues from 92 patients who received curative resection, higher hepcidin tissue expression was associated with significantly poorer overall survival and was correlated with more advanced PDAC stage and vascular invasion (39). Local upregulation expression of hepcidin has also been reported in other tumor tissues (40–45). The increased hepcidin expression by cancer cells is hypothesized to be an adaptation that affects iron export and contributes to iron retention in tumor cells, helping them survive and proliferate (45). Furthermore, TFR1 was highly expressed in PDAC cells and was necessary for pancreatic cell proliferation and tumorigenesis through mitochondrial respiration and ROS formation (46). In a study that included tissues from 96 PDAC patients, iron content was higher in tumor tissue compared with adjacent normal tissue, and patients with high tumor TFR1 expression had significantly worse overall and relapse-free survival compared with those with negative/low/medium expression (47). More research is needed to understand the role of the other genes selected in our pathway analysis in PDAC.
In our exploratory analyses, the hepcidin-regulating iron metabolism gene pathway overall was associated with all the iron status biomarkers, and the sARTP-selected SNPs for PDAC were associated with the iron status biomarkers and/or hematology traits. The sARTP-selected SNPs for TFR1 were positively associated with PDAC (Table 2) and with ferritin, suggesting higher iron stores and greater PDAC risk (Supplemental Figure 2). Alleles in the sARTP-selected SNPs for TFR2 (rs62482223-C, rs56328569-C) were inversely associated with PDAC and also inversely associated with serum iron, transferrin saturation, consistent with lower iron status contributing to reduced PDAC risk. The sARTP-selected HAMP SNP rs10421768 (G compared with the C allele) associated with increased PDAC in our study is in the 5′ flanking region of the HAMP gene encoding hepcidin and has been associated with higher liver (48) and cardiac tissue (49) and higher ferritin concentrations (48) in β-thalassemia patients. One study that included only men to avoid effects of physiologically lower serum ferritin levels in women due to menstruation reported rs10421768-G was associated with lower serum ferritin concentrations (50). Our analysis of participants in the Genetics of Iron Status Consortium and another study of 244 healthy individuals (51) did not demonstrate associations between rs10421768-G and serum iron, ferritin, or transferrin concentrations; however, neither study was able to stratify associations by sex. Although iron plays a key role in hematologic traits and some associations between the sARTP-selected SNPs with the hematologic traits were identified, it is unclear how this may be related to PDAC and thus requires further investigation.
Some sARTP-selected SNPs located in HJV and HAMPact as eQTLs in normal pancreas tissue, adding to the biologic plausibility of our findings. Alleles in 4 correlated sARTP-selected SNPs (rs10419959-A, rs12981457-T, rs10421768-G, rs10424619-T) in HAMP were associated with expression of UFS2 in normal pancreas tissue. USF2plays a role in hepcidin regulation (52). Compared with wild type, Usf2–/– knockout mice had progressive iron accumulation up to 20-fold higher in the pancreas and 10-fold higher in the liver between 60 and 100 d after birth (52) and without hepcidin expression in the liver (51). We also observed eQTL signals in NBPF10 for 2 correlated sARTP-selected HJV SNPs (rs6424377, rs10910810). Although NBPF10’s function is unknown, it is located in close proximity to HJV. Rare pathogenic mutations in both HJV and HAMP are known to play a role in juvenile hemochromatosis.
The strengths of our study include the large numbers of PDAC cases and controls and the statistical pathway approach using GWAS summary statistics, which allows for detection of the accumulative effect of multiple PDAC-associated variants within genes and surroundings regions in the hepcidin-regulating iron metabolism pathway. Our approach provides the opportunity to identify susceptibility related to iron metabolism beyond that from dietary iron intake. The exploratory analysis with iron status and hematologic traits, as well as the eQTL and functional annotations, adds to the biologic plausibility of the associations we observed. Our study also has limitations. The gene regions included in our analyses may not functionally influence the hepcidin pathway gene of interest but may be of functional importance for other nearby genes not part of the hepcidin pathway. Additional functional studies are needed to verify that observed associations represent a biologic relationship between the hepcidin-regulating iron metabolism pathway and PDAC. We do not have sufficient dietary data or power to evaluate pathway subgroup analyses stratified by iron or meat intake. The strongest pathway association was observed in the full set of cases and controls. Extreme care is warranted in interpretation of associations in subgroups (i.e., total meats and iron intakes, women, diagnosed at age >60 y, and BMI >30) because subgroup associations are known to be unreliable and underpowered: Spurious subgroup effects could be identified when none exist. Our analysis does not include less common pathogenic variants known to play a role in hereditary hemochromatosis; however, it does utilize common variants within the same genes. Individuals in our study were of European ancestry and most were aged >50 y; therefore, our findings may not be generalizable to other ancestral groups or younger individuals.
In conclusion, our results using common variants from GWAS support the hypothesis that the hepcidin-regulating iron metabolism pathway based on genes involved in iron sensing and regulation of dietary iron absorption is associated with PDAC. Further epidemiologic and experimental studies are needed to confirm our findings and to better understand the biologic mechanisms contributing to PDAC etiology.
Supplementary Material
Acknowledgments
We thank Barry Graubard and Tomas Ganz for their assistance in data interpretation. We acknowledge Kathy Helzlsouer for her contribution to CLUE2 and PanScan. We also acknowledge the contribution of Irene Orlow to this analysis. The cooperation of 30 Connecticut hospitals, including Stamford Hospital, in allowing patient access is gratefully acknowledged. We acknowledge the contribution of Frederike Dijk and Oliver Busch (Academic Medical Center, Amsterdam, Netherlands). Assistance with genotype data quality control was provided by Cecelia Laurie and Cathy Laurie at the University of Washington Genetic Analysis Center. We acknowledge the State of Maryland, the Maryland Cigarette Restitution Fund, and the National Program of Cancer Registries of the CDC for funding the collection and availability of the Maryland Cancer Registry data. We thank the Women's Health Initiative investigators and staff (a full listing of the investigators can be found at http://www.whi.org/researchers/Documents%20%20Write%20a%20Paper/WHI%20Investigator%20Long%20List.pdf).
The authors’ responsibilities were as follows—SJS and RZS-S: designed the research; KY: provided statistical methods and advice on their application; WW: performed statistical analysis; BB, AAA, LEB-F., PMB, EJD, MD, SG, GGG, PJG, CK, LLM, REN, X-OS, SKVDE, KV, WZ, DA, GA, AB, SIB, LKB, PB, BBdM, JEB, SJC, CCC, LF, CSF, JMG, MGG, TH, PH, MMH, IH, EAH, RIH, VJ, RCK, I-ML, NM, RLM, ALO, UP, MP, NR, MBS, HDS, DTS, IMT, JW-W, NW, E White, LRW, HY, AZ-J., PK, DL, GMP, BMW, HAR, LTA, APK, and RZS-S: provided essential reagents or provided essential materials; SJ-S, KY, and RZS-S: wrote the manuscript; SJ-S and RZS-S: had primary responsibility for final content; and all authors: contributed substantive interpretation of and editorial comments on the manuscript drafts and reviewed, read, and approved the final manuscript. The authors report no conflicts of interest.
Notes
This work was supported by the Intramural Research Program, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health and RO1CA154823. The International Agency for Research on Cancer/Central Europe study was supported by grant R03CA123546-02 from the NCI–NIH and grants from the Ministry of Health of the Czech Republic (NR 9029–4/2006, NR9422-3, NR9998-3, and MH CZ-DRO-MMCI 00209805). The work at Johns Hopkins University was supported by NCI grants P50CA062924, R01CA97075, R01CA154382, and U01CA247283. Additional support was provided by the Lustgarten Foundation, Susan Wojcicki and Dennis Troper, and the Sol Goldman Pancreas Cancer Research Center. The Mayo Clinic Biospecimen Resource for Pancreas Research study is supported by the Mayo Clinic SPORE in Pancreatic Cancer (P50 CA102701). The MD Anderson case–control study was supported by NIH–NCI grant R01CA098380. The Pancreatic Cancer Cohort II study was partially supported by a supplemental grant to CA098380. The Multiethnic Cohort study was supported by NCI grant U01CA164973. The Memorial Sloan Kettering Cancer Center Pancreatic Tumor Registry is supported by P30CA008748, the Geoffrey Beene Foundation, the Arnold and Arlene Goldstein Family Foundation, and the Society of MSKCC. The PACIFIC study was supported by grant RO1CA102765 from the Kaiser Permanente and Group Health Cooperative. The Queensland Pancreatic Cancer Study was supported by grant 442302 from the National Health and Medical Research Council of Australia (NHMRC). REN is supported by a NHMRC Senior Research Fellowship (1060183). The University of California, San Francisco, pancreas study was supported by NIH–NCI grants R01CA1009767, R01CA109767-S1, and R0CA059706 and by the Joan Rombauer Pancreatic Cancer Fund. Collection of cancer incidence data was supported by the California Department of Public Health as part of the statewide cancer reporting program; the NCI's SEER Program under contract HSN261201000140C awarded to CPIC; and the CDC's National Program of Cancer Registries under agreement U58DP003862-01 awarded to the California Department of Public Health. The Yale pancreas cancer study is supported by NCI–NIH grant 5R01CA098870. The Connecticut Pancreas Cancer Study was approved by the State of Connecticut Department of Public Health Human Investigation Committee. Certain data used in that study were obtained from the Connecticut Tumor Registry in the Connecticut Department of Public Health. The authors assume full responsibility for analyses and interpretation of these data. Studies included in PANDoRA were partly funded by the Czech Science Foundation (P301/12/1734); the Internal Grant Agency of the Czech Ministry of Health (IGA NT 13 263); the Baden–Württemberg State Ministry of Research, Science and Arts (H. Brenner); the Heidelberger EPZ-Pancobank (M. W. Büchler and team: T. Hackert, N. A. Giese, Ch. Tjaden, E. Soyka, M. Meinhardt; Heidelberger Stiftung Chirurgie and BMBF grant 01GS08114); the BMBH (P. Schirmacher; BMBF grant 01EY1101); the “5 × 1000” voluntary contribution of the Italian Government; the Italian Ministry of Health (RC1203GA57, RC1303GA53, RC1303GA54, and RC1303GA50); the Italian Association for Research on Cancer (A. Scarpa; AIRC 12182); the Italian Ministry of Research (A. Scarpa; FIRB-RBAP10AHJB); the Italian FIMP-Ministry of Health (A. Scarpa; 12 CUP_J33G13000210001); and the National Institute for Health Research Liverpool Pancreas Biomedical Research Unit, UK. The American Cancer Society funds the creation, maintenance, and updating of the Cancer Prevention Study II cohort. Cancer incidence data for CLUE were provided by the Maryland Cancer Registry, Center for Cancer Surveillance and Control, Department of Health and Mental Hygiene (http://phpa.dhmh.maryland.gov/cancer). The Melbourne Collaborative Cohort Study (MCCS) cohort recruitment was funded by VicHealth and Cancer Council Victoria. The MCCS was further augmented by Australian National Health and Medical Research Council grants 209057, 396414, and 1074383 and by infrastructure provided by Cancer Council Victoria. Cases and their vital status were ascertained through the Victorian Cancer Registry and the Australian Institute of Health and Welfare, including the National Death Index and the Australian Cancer Database. The New York University study (AZJ and AAA) was funded by NIH grants R01CA098661 and UM1CA182934 and center grants P30CA016087 and P30ES000260. The PANKRAS II study in Spain was supported by research grants from Instituto de Salud Carlos III-FEDER, Spain: Fondo de Investigaciones Sanitarias (FIS: PI13/00082 and PI15/01573) and Red Temática de Investigación Cooperativa en Cáncer, Spain (RD12/0036/0050); and European Cooperation in Science and Technology (COST Action BM1204: EU_Pancreas), Ministerio de Ciencia y Tecnología (CICYT SAF 2000–0097), Fondo de Investigación Sanitaria (95/0017), Madrid, Spain; Generalitat de Catalunya (CIRIT - SGR); “Red temática de investigación cooperativa de centros en Cáncer” (C03/10), “Red temática de investigación cooperativa de centros en Epidemiología y salud pública” (C03/09), and CIBER de Epidemiología (CIBERESP), Madrid, Spain. The Physicians’ Health Study was supported by research grants CA097193, CA34944, CA40360, HL26490, and HL34595 from the NIH. The Women's Health Study was supported by research grants CA182913, CA047988, HL043851, HL080467, and HL099355 from the NIH. The Health Professionals Follow-up Study is supported by NIH grant UM1CA167552 from the NCI. The Nurses’ Health Study is supported by NIH grants UM1 CA186107, P01CA87969, and R01CA49449 from the NCI. Additional support was provided from the Hale Center for Pancreatic Cancer Research, U01CA21017 from the NCI, the US Department of Defense (CA130288), Lustgarten Foundation, Pancreatic Cancer Action Network, Noble Effort Fund, Peter R. Leavitt Family Fund, Wexler Family Fund, and Promises for Purple to BMW. Fundings for EJ Duell included Instituto de Salud Carlos III, cofunded by FEDER funds—a way to build Europe—(FIS: PI15/00659 and PI18-00192); Fundació La Marató de TV3 (20192-30–); Agency for Management of University and Research Grants (AGAUR) of the Catalan Government grant 2017SGR723; and the Spanish Association Against Cancer Scientific Foundation. The Shanghai Men's Health Study is supported by NIH grant UM1CA173640. The Shanghai Women's Health Study is supported by NIH grant UM1CA182910. The Women's Health Initiative (WHI) program is funded by the National Heart, Lung, and Blood Institute, NIH, US Department of Health and Human Services through contracts HHSN268201600018C, HHSN268201600001C, HHSN268201600002C, HHSN268201600003C, and HHSN268201600004C. The SELECT study is supported by NIH grant U10CA37429 (CDB) and UM1CA182883 (C. M. Tangen and IMT). This study utilized the high-performance computational capabilities of the Biowulf Linux cluster at the NIH ( https://hpc.nih.gov/systems/ ). The Genotype-Tissue Expression (GTEx) Project was supported by the Common Fund of the Office of the Director of the NIH and by NCI; the National Human Genome Research Institute; the National Heart, Lung, and Blood Institute; the National Institute on Drug Abuse; the National Institute of Mental Health; and the National Institute of Neurological Disorders and Stroke. The data used for the analyses described in this article were obtained from the tissue data from the GTEx Portal on June 2019 and July 2020.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Where authors are identified as personnel of the International Agency for Research on Cancer/WHO, the authors alone are responsible for the views expressed in this article and they do not necessarily represent the decisions, policy, or views of the International Agency for Research on Cancer/WHO.
Supplemental Figures 1 and 2 and Supplemental Tables 1–6 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/ajcn/.
Abbreviations used: BMP, bone morphogenetic protein; eQTL, expression quantitative trait loci;FTH1, ferritin heavy chain 1; FTL, ferritin light chain; GTEx, Genotype-Tissue Expression; GWAS, genome-wide association study;HAMP, hepcidin; HFE, homeostatic iron regulator; HJV, hemojuvelin; LD, linkage disequilibrium; NCI, National Cancer Institute; NRF2, nuclear factor erythroid 2-related factor 2; PanC4, Pancreatic Cancer Case–Control Consortium; PanScan, Pancreatic Cancer Cohort Consortium; PDAC, pancreatic ductal adenocarcinoma; RDW, red blood cell distribution width; ROS, reactive oxygen species; sARTP, summary data-based adaptive rank truncated product; SLC40A1, ferroportin 1; SNP, single nucleotide polymorphism; TFR, transferrin receptor.
Contributor Information
Sachelly Julián-Serrano, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Fangcheng Yuan, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
William Wheeler, Information Management Services, Silver Spring, MD, USA.
Beben Benyamin, Australian Centre for Precision Health, Allied Health and Human Performance, University of South Australia, Adelaide, Australia; South Australian Health and Medical Research Institute, Adelaide, Australia.
Mitchell J Machiela, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Alan A Arslan, Department of Obstetrics and Gynecology, New York University School of Medicine, New York, NY, USA.
Laura E Beane-Freeman, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Paige M Bracci, Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, USA.
Eric J Duell, Unit of Biomarkers and Susceptibility, Oncology Data Analytics Program, Catalan Institute of Oncology, L'Hospitalet de Llobregat, Barcelona, Spain; Colorectal Cancer Group, ONCOBELL Program, Bellvitge Biomedical Research Institute (IDIBELL), L'Hospitalet de Llobregat, Barcelona, Spain; Consortium for Biomedical Research in Epidemiology and Public Health (CIBERESP), Madrid, Spain.
Mengmeng Du, Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, USA.
Steven Gallinger, Lunenfeld–Tanenbaum Research Institute, Sinai Health System, Toronto, Canada.
Graham G Giles, Cancer Epidemiology Division, Cancer Council Victoria, Melbourne, Australia; Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Australia; Precision Medicine, School of Clinical Sciences at Monash Health, Monash University, Clayton, Australia.
Phyllis J Goodman, SWOG Statistical Center, Fred Hutchinson Cancer Research Center, Seattle, WA, USA.
Charles Kooperberg, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA, USA.
Loic Le Marchand, Department of Epidemiology, University of Hawaii Cancer Center, Honolulu, HI, USA.
Rachel E Neale, Population Health Department, QIMR Berghofer Medical Research Institute, Brisbane, Australia.
Xiao-Ou Shu, Division of Epidemiology, Department of Medicine, Vanderbilt Epidemiology Center, Vanderbilt–Ingram Cancer Center, Vanderbilt University School of Medicine, Nashville, TN, USA.
Stephen K Van Den Eeden, Division of Research, Kaiser Permanente Northern California, Oakland, CA, USA.
Kala Visvanathan, Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA.
Wei Zheng, Division of Epidemiology, Department of Medicine, Vanderbilt Epidemiology Center, Vanderbilt–Ingram Cancer Center, Vanderbilt University School of Medicine, Nashville, TN, USA.
Demetrius Albanes, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Gabriella Andreotti, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Eva Ardanaz, Navarra Public Health Institute, Pamplona, Spain; IdiSNA, Navarra Institute for Health Research, Pamplona, Spain; CIBER Epidemiología y Salud Pública (CIBERESP), Barcelona, Spain.
Ana Babic, Department of Medical Oncology, Dana–Farber Cancer Institute, Boston, MA, USA.
Sonja I Berndt, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Lauren K Brais, Department of Medical Oncology, Dana–Farber Cancer Institute, Boston, MA, USA.
Paul Brennan, International Agency for Research on Cancer (IARC), Lyon, France.
Bas Bueno-de-Mesquita, Department for Determinants of Chronic Diseases, National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands.
Julie E Buring, Division of Preventive Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, MA, USA.
Stephen J Chanock, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Erica J Childs, Department of Oncology, The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore, MD, USA.
Charles C Chung, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Eleonora Fabiánová, Specialized Institute of Hygiene and Epidemiology, Banska Bystrica, Slovakia.
Lenka Foretová, Department of Cancer Epidemiology and Genetics, Masaryk Memorial Cancer Institute, Brno, Czech Republic.
Charles S Fuchs, Yale Cancer Center and Smilow Cancer Hospital, New Haven, CT, USA.
J Michael Gaziano, Division of Aging, Brigham and Women's Hospital, Boston, MA, USA.
Manuel Gentiluomo, Department of Biology, University of Pisa, Italy; Genomic Epidemiology Group, German Cancer Research Center, (DKFZ), Heidelberg, Germany.
Edward L Giovannucci, Department of Medical Oncology, Dana–Farber Cancer Institute, Boston, MA, USA.
Michael G Goggins, Department of Pathology, Sol Goldman Pancreatic Cancer Research Center, Johns Hopkins School of Medicine, Baltimore, MD, USA.
Thilo Hackert, Department of General, Visceral and Transplantation Surgery, Heidelberg University Hospital, Heidelberg, Germany.
Patricia Hartge, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Manal M Hassan, Department of Epidemiology, University of Texas MD Anderson Cancer Center, Houston, TX, USA.
Ivana Holcátová, Institute of Public Health and Preventive Medicine, Second Faculty of Medicine, Charles University, Prague, Czech Republic.
Elizabeth A Holly, Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, USA.
Rayjean I Hung, Lunenfeld–Tanenbaum Research Institute, Sinai Health System, Toronto, Canada.
Vladimir Janout, Faculty of Health Sciences, University of Olomouc, Olomouc, Czech Republic.
Robert C Kurtz, Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, NY, USA.
I-Min Lee, Division of Preventive Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, MA, USA; Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA, USA.
Núria Malats, Genetic and Molecular Epidemiology Group, Spanish National Cancer Research Centre (CNIO), Madrid, Spain.
David McKean, Department of Oncology, The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore, MD, USA.
Roger L Milne, Cancer Epidemiology Division, Cancer Council Victoria, Melbourne, Australia; Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Australia; Precision Medicine, School of Clinical Sciences at Monash Health, Monash University, Clayton, Australia.
Christina C Newton, Department of Population Science, American Cancer Society, Atlanta, GA, USA.
Ann L Oberg, Department of Health Sciences Research, Mayo Clinic College of Medicine, Rochester, MN, USA.
Sandra Perdomo, International Agency for Research on Cancer (IARC), Lyon, France.
Ulrike Peters, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA, USA.
Miquel Porta, Hospital del Mar Institute of Medical Research (IMIM), Universitat Autònoma de Barcelona, Barcelona, Spain.
Nathaniel Rothman, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Matthias B Schulze, Department of Molecular Epidemiology, German Institute of Human Nutrition Potsdam-Rehbruecke, Nuthetal, Germany; Institute of Nutritional Science, University of Potsdam, Nuthetal, Germany.
Howard D Sesso, Division of Preventive Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, MA, USA.
Debra T Silverman, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Ian M Thompson, CHRISTUS Santa Rosa Hospital–Medical Center, San Antonio, TX, USA.
Jean Wactawski-Wende, Department of Epidemiology and Environmental Health, University at Buffalo, Buffalo, NY, USA.
Elisabete Weiderpass, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA, USA.
Nicolas Wenstzensen, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Emily White, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA, USA.
Lynne R Wilkens, Department of Epidemiology, University of Hawaii Cancer Center, Honolulu, HI, USA.
Herbert Yu, Department of Epidemiology, University of Hawaii Cancer Center, Honolulu, HI, USA.
Anne Zeleniuch-Jacquotte, Department of Population Health and Perlmutter Cancer Center, New York University School of Medicine, New York, NY, USA.
Jun Zhong, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Peter Kraft, Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA, USA; Department of Biostatistics, Harvard T.H. Chan School of Public Health, Boston, MA, USA.
Dounghui Li, Department of Gastrointestinal Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, USA.
Peter T Campbell, Department of Population Science, American Cancer Society, Atlanta, GA, USA.
Gloria M Petersen, Department of Health Sciences Research, Mayo Clinic College of Medicine, Rochester, MN, USA.
Brian M Wolpin, Department of Medical Oncology, Dana–Farber Cancer Institute, Boston, MA, USA.
Harvey A Risch, Department of Chronic Disease Epidemiology, Yale School of Public Health, New Haven, CT, USA.
Laufey T Amundadottir, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Alison P Klein, Department of Oncology, The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore, MD, USA; Department of Pathology, Sol Goldman Pancreatic Cancer Research Center, Johns Hopkins School of Medicine, Baltimore, MD, USA.
Kai Yu, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Rachael Z Stolzenberg-Solomon, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.
Data Availability
Data described in the manuscript and code book are available on dbGAP. Biomedical research scientists from recognized research institutions can request the data from dbGAP as bona fide researchers.
References
- 1.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin. 2020;70(1):7–30. [DOI] [PubMed] [Google Scholar]
- 2.Ryan DP, Hong TS, Bardeesy N. Pancreatic adenocarcinoma. N Engl J Med. 2014;371(11):1039–49. [DOI] [PubMed] [Google Scholar]
- 3.Chen F, Childs EJ, Mocci E, Bracci P, Gallinger S, Li D, Neale RE, Olson SH, Scelo G, Bamlet WRet al. Analysis of heritability and genetic architecture of pancreatic cancer: a PanC4 study. Cancer Epidemiol Biomarkers Prev. 2019;28(7):1238–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Zhao Z, Yin Z, Pu Z, Zhao Q. Association between consumption of red and processed meat and pancreatic cancer risk: a systematic review and meta-analysis. Clin Gastroenterol Hepatol. 2017;15(4):486–493.e10. [DOI] [PubMed] [Google Scholar]
- 5.Bouvard V, Loomis D, Guyton KZ, Grosse Y, Ghissassi FE, Benbrahim-Tallaa L, Guha N, Mattock H, Straif K; International Agency for Research on Cancer Monograph Working Group . Carcinogenicity of consumption of red and processed meat. Lancet Oncol. 2015;16(16):1599–600. [DOI] [PubMed] [Google Scholar]
- 6.Taunk P, Hecht E, Stolzenberg-Solomon R. Are meat and heme iron intake associated with pancreatic cancer? Results from the NIH–AARP diet and health cohort. Int J Cancer. 2016;138(9):2172–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.World Cancer Research Fund/American Institute for Cancer Research . Diet, nutrition, physical activity and pancreatic cancer: Continuous Update Project expert report 2018. London: World Cancer Research Fund/American Institute for Cancer Research; 2018. [Google Scholar]
- 8.Hooda J, Shah A, Zhang L. Heme, an essential nutrient from dietary proteins, critically impacts diverse physiological and pathological processes. Nutrients. 2014;6(3):1080–102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Friedman GD, van den Eeden SK. Risk factors for pancreatic cancer: an exploratory study. Int J Epidemiol. 1993;22(1):30–7. [DOI] [PubMed] [Google Scholar]
- 10.Torti SV, Torti FM. Iron and cancer: more ore to be mined. Nat Rev Cancer. 2013;13(5):342–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Gaur A, Collins H, Wulaningsih W, Holmberg L, Garmo H, Hammar N, Walldius G, Jungner I, Van Hemelrijck M. Iron metabolism and risk of cancer in the Swedish AMORIS study. Cancer Causes Control. 2013;24(7):1393–402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Backe MB, Moen IW, Ellervik C, Hansen JB, Mandrup-Poulsen T. Iron regulation of pancreatic β-cell functions and oxidative stress. Annu Rev Nutr. 2016;36:1, 241–73. [DOI] [PubMed] [Google Scholar]
- 13.Rahier J, Loozen S, Goebbels RM, Abrahem M. The haemochromatotic human pancreas: a quantitative immunohistochemical and ultrastructural study. Diabetologia. 1987;30(1):5–12. [DOI] [PubMed] [Google Scholar]
- 14.Ganz T. Systemic iron homeostasis. Physiol Rev. 2013;93(4):1721–41. [DOI] [PubMed] [Google Scholar]
- 15.Ganz T, Nemeth E. Hepcidin and iron homeostasis. Biochim Biophys Acta. 2012;1823(9):1434–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ginzburg YZ. Hepcidin–ferroportin axis in health and disease. Vitam Horm. 2019;110:17–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Drakesmith H, Prentice AM. Hepcidin and the iron–infection axis. Science. 2012;338(6108):768–72. [DOI] [PubMed] [Google Scholar]
- 18.Brissot P, Pietrangelo A, Adams PC, de Graaff B, McLaren CE. Haemochromatosis. Nat Rev Dis Primers. 2018;4:18016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Hart PA, Bellin MD, Andersen DK, Bradley D, Cruz-Monserrate Z, Forsmark CE, Goodarzi MO, Habtezion A, Korc M, Kudva YCet al. Type 3c (pancreatogenic) diabetes mellitus secondary to chronic pancreatitis and pancreatic cancer. Lancet Gastroenterol Hepatol. 2016;1(3):226–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Amundadottir LT, Kraft P, Stolzenberg-Solomon RZ, Fuchs CS, Petersen GM, Arslan AA, Bueno-de-Mesquita H, Gross M, Helzlsouer K, Jacobs Eet al. Genome-wide association study identifies variants in the ABO locus associated with susceptibility to pancreatic cancer. Nat Genet. 2009;41(9):986–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Petersen G, Amundadottir L, Fuchs CS, Kraft P, Stolzenberg-Solomon RZ, Jacobs K, Arslan AA, Bueno-de-Mesquita H, Gallinger S, Gross Met al. A genome-wide association study identifies pancreatic cancer susceptibility loci on chromosomes 13q22.1, 1q32.1 and 5p15.33. Nat Genet. 2010;42(3):224–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Wolpin B, Rizzato C, Kraft P, Kooperberg C, Petersen G, Wang Z, Arslan AA, Beane-Freeman L, Bracci P, Buring Jet al. Genome-wide association study identifies multiple susceptibility loci for pancreatic cancer. Nat Genet. 2014;46(9):994–1000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Childs E, Mocci E, Campa D, Bracci P, Gallinger S, Goggins M, Li D, Neale RE, Olson SH, Scelo Get al. Common variation at 2p13.3, 3q29, 7p13 and 17q25.1 associated with susceptibility to pancreatic cancer. Nat Genet. 2015;47(8):911–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Klein AP, Wolpin BM, Risch HA, Stolzenberg-Solomon RZ, Mocci E, Zhang M, Canzian F, Childs EJ, Hoskins JW, Jermusyk Aet al. Genome-wide meta-analysis identifies five new susceptibility loci for pancreatic cancer. Nat Commun. 2018;9(1):556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Howie BN, Donnelly P, Marchini J. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet. 2009;5(6):e1000529. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wang CY, Babitt JL. Liver iron sensing and body iron homeostasis. Blood. 2019;133(1):18–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hentze MW, Muckenthaler MU, Galy B, Camaschella C. Two to tango: regulation of mammalian iron metabolism. Cell. 2010;142(1):24–38. [DOI] [PubMed] [Google Scholar]
- 28.Muckenthaler MU, Rivella S, Hentze MW, Galy B. A red carpet for iron metabolism. Cell. 2017;168(3):344–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zhang H, Wheeler W, Hyland PL, Yang Y, Shi J, Chatterjee N, Yu K. A powerful procedure for pathway-based meta-analysis using summary statistics identifies 43 pathways associated with type II diabetes in European populations. PLoS Genet. 2016;12(6):e1006122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Consortium GT. The Genotype-Tissue Expression (GTEx) project. Nat Genet. 2013;45(6):580–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ward LD, Kellis M. HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Res. 2012;40(D1):D930–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Boyle AP, Hong EL, Hariharan M, Cheng Y, Schaub MA, Kasowski M, Karczewski KJ, Park J, Hitz BC, Weng Set al. Annotation of functional variation in personal genomes using RegulomeDB. Genome Res. 2012;22(9):1790–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Machiela MJ, Chanock SJ. LDlink: a Web-based application for exploring population-specific haplotype structure and linking correlated alleles of possible functional variants. Bioinformatics. 2015;31(21):3555–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Benyamin B, Esko T, Ried JS, Radhakrishnan A, Vermeulen SH, Traglia M, Gogele M, Anderson D, Broer L, Podmore Cet al. Novel loci affecting iron homeostasis and their effects in individuals at risk for hemochromatosis. Nat Commun. 2014;5:4926. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Canela-Xandri O, Rawlik K, Tenesa A. An atlas of genetic associations in UK Biobank. Nat Genet. 2018;50(11):1593–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Sangkhae V, Nemeth E. Regulation of the iron homeostatic hormone hepcidin. Adv Nutr. 2017;8(1):126–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Zhang M, Xiong H, Fang L, Lu W, Wu X, Wang YQ, Cai ZM, Wu S. Meta-analysis of the association between H63D and C282Y polymorphisms in HFE and cancer risk. Asian Pac J Cancer Prev. 2015;16(11):4633–9. [DOI] [PubMed] [Google Scholar]
- 38.Zhao Z, Li C, Hu M, Li J, Liu R. Plasma ferritin levels, HFE polymorphisms, and risk of pancreatic cancer among Chinese Han population. Tumour Biol. 2014;35(8):7629–33. [DOI] [PubMed] [Google Scholar]
- 39.Toshiyama R, Konno M, Eguchi H, Asai A, Noda T, Koseki J, Asukai K, Ohashi T, Matsushita K, Iwagami Yet al. Association of iron metabolic enzyme hepcidin expression levels with the prognosis of patients with pancreatic cancer. Oncol Lett. 2018;15(5):8125–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Pinnix ZK, Miller LD, Wang W, D'Agostino R Jr, Kute T, Willingham MC, Hatcher H, Tesfay L, Sui G, Di Xet al. Ferroportin and iron regulation in breast cancer progression and prognosis. Sci Transl Med. 2010;2(43):43ra56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Xiang-Tao P. Expression of hepcidin and neogenin in colorectal cancer. Open Medicine. 2017;12:184–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Ward DG, Roberts K, Brookes MJ, Joy H, Martin A, Ismail T, Spychal R, Iqbal T, Tselepis C. Increased hepcidin expression in colorectal carcinogenesis. World J Gastroenterol. 2008;14(9):1339–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Tesfay L, Clausen KA, Kim JW, Hegde P, Wang X, Miller LD, Deng Z, Blanchette N, Arvedson T, Miranti CKet al. Hepcidin regulation in prostate and its disruption in prostate cancer. Cancer Res. 2015;75(11):2254–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Wang F, Liu A, Bai R, Zhang B, Jin Y, Guo W, Li Y, Gao J, Liu L. Hepcidin and iron metabolism in the pathogenesis of prostate cancer. J BUON. 2017;22(5):1328–32. [PubMed] [Google Scholar]
- 45.Vela D, Vela-Gaxha Z. Differential regulation of hepcidin in cancer and non-cancer tissues and its clinical implications. Exp Mol Med. 2018;50(2):e436. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Jeong SM, Hwang S, Seong RH. Transferrin receptor regulates pancreatic cancer growth by modulating mitochondrial respiration and ROS generation. Biochem Biophys Res Commun. 2016;471(3):373–9. [DOI] [PubMed] [Google Scholar]
- 47.Lang J, Zhao X, Wang X, Zhao Y, Li Y, Zhao R, Cheng K, Li Y, Han X, Zheng Xet al. Targeted co-delivery of the iron chelator deferoxamine and a HIF1α inhibitor impairs pancreatic tumor growth. ACS Nano. 2019;13(2):2176–89. [DOI] [PubMed] [Google Scholar]
- 48.Andreani M, Radio FC, Testi M, De Bernardo C, Troiano M, Majore S, Bertucci P, Polchi P, Rosati R, Grammatico P. Association of hepcidin promoter c.-582 A>G variant and iron overload in thalassemia major. Haematologica. 2009;94(9):1293–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Sukiennicki GM, Marciniak W, Muszynska M, Baszuk P, Gupta S, Bialkowska K, Jaworska-Bieniek K, Durda K, Lener M, Pietrzak Set al. Iron levels, genes involved in iron metabolism and antioxidative processes and lung cancer incidence. PLoS One. 2019;14(1):e0208610. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Bruno F, Bonalumi S, Camaschella C, Ferrari M, Cremonesi L. The -582A>G variant of the HAMP promoter is not associated with high serum ferritin levels in normal subjects. Haematologica. 2010;95(5):849–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Nicolas G, Bennoun M, Devaux I, Beaumont C, Grandchamp B, Kahn A, Vaulont S. Lack of hepcidin gene expression and severe tissue iron overload in upstream stimulatory factor 2 (USF2) knockout mice. Proc Natl Acad Sci USA. 2001;98(15):8780–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Bayele HK, McArdle H, Srai SK. Cis and trans regulation of hepcidin expression by upstream stimulatory factor. Blood. 2006;108(13):4237–45. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data described in the manuscript and code book are available on dbGAP. Biomedical research scientists from recognized research institutions can request the data from dbGAP as bona fide researchers.