Abstract
Background: Recent reports support an association between chronic inflammation and progression to pancreatic cancer (PC).
Methods: This case–control, candidate gene association study evaluated 1,354 Caucasian patients with pancreatic ductal adenocarcinoma and 1,189 healthy Caucasian controls. We genotyped 1,538 single nucleotide polymorphism (SNP) in 102 genes from inflammatory pathways involving NF-κB. Primary tests of association assumed a multiplicative (log-additive) genotype effect; secondary analyses examined dominant, additive, and recessive SNP effects.
Results: After adjusting for known risk factors for PC, single SNP analysis revealed an association between four SNPs in NOS1 and one in the CD101 gene with PC risk. These results, however, were not replicated in a PC case–control and cohort population.
Conclusion:NOS1 and CD101 may be associated with a risk of PC; however, these findings did not replicate in other PC populations. Future research is needed into the possible role of NOS1 and CD101 for PC.
Impact: This research shows a lack of association between genetic variation in 102 inflammation-related genes and PC. Future research is needed into the possible role of other inflammation-related genes and PC risk. Cancer Epidemiol Biomarkers Prev; 20(6); 1251–4. ©2011 AACR.
Introduction
Chronic inflammation has been recognized as a contributing factor in the development of a subset of highly lethal pancreatic adenocarcinomas (PC; ref. 1). Studies linking chronic inflammation to PC report an upregulation of cyclooxygenase (COX; ref. 2) and nitric oxide (NO) gene expression (3), important mediators of inflammation in PC tissue specimens. The trigger for upregulation of COX and NO is the transcription factor, NF-κB. NF-κB activates these genes and others that are involved in inflammation and apoptosis and seems to promote pancreatic cell growth by inhibition of apoptosis (4). We hypothesized that genetic polymorphisms in inflammation-related genes involving NF-κB–related inflammatory pathways are associated with risk of PC.
Methods
PC patients were recruited prospectively to a PC research registry using processes approved by the Mayo Clinic Institutional Review Board. Consenting patients completed a comprehensive questionnaire, donated a blood sample, and provided medical release for study of archival tissue.
Subjects
Cases were patients with histologically proven pancreatic ductal adenocarcinoma evaluated at Mayo Clinic from 2000 to 2008. Controls were healthy, clinic-based patients without a personal history of cancer (except nonmelanoma skin cancer) and were frequency matched to cases for age (±5 years), sex, race, and geographic region of residence (5).
Candidate gene and single nucleotide polymorphism selection
Candidate genes were selected using a combination of literature review and the bioinformatics tools Ingenuity Systems and MetaCore from GeneGo, Inc. to identify genes involved in the pathogenesis of PC that are in the NF-κB–related inflammatory pathways. Genotyping and Quality Control measures were followed as reported previously (5). Call rates for single nucleotide polymorphisms (SNP) were 99.2% and 97.3%. Eighty-four of 1,538 SNPs failed to amplify.
Statistical analysis
Unconditional logistic regression analysis was used to estimate ORs and 95% CIs for the risk of PC with each SNP. The regression model included covariates of age, sex, smoking status, body mass index (kg/m2), family history of PC, and preexisting diabetes (>2 years). Primary tests of association assumed a multiplicative (log-additive) genotype effect, equivalent to the Armitage test for trend. To account for multiple testing, SNPs were deemed significantly associated with PC if P < 0.001.
Validation studies
Genotyping data from PanScan, the Pancreatic Cancer Cohort Consortium, and Pancreatic Cancer Case-Control Consortium (6, 7) were obtained from dbGap (8). Imputation in non-Mayo samples was accomplished with MACH software (9) using HapMap. The imputed allele dosage was used to validate SNPs of interest. All analyses were unadjusted.
Results
Demographics for all subjects are reported in Table 1. Single SNP analysis revealed an association between 4 SNPs in NOS1 and 1 SNP in the CD101 gene and the risk of PC (Table 2). An additional 14 and 4 imputed SNPs with values of P = 5E−4 to 8E−4 were observed in CD101 and NOS1, respectively (data not shown). The significant SNPs failed confirmation in the validation set using both PanScan data sets.
. | Control (N = 1,171) . | Case (N = 1,329) . | Total (N = 2,500) . | P . |
---|---|---|---|---|
Sex | ||||
Female | 574 (49%) | 564 (42.4%) | 1,138 (45.5%) | 0.0010 |
Male | 597 (51%) | 765 (57.6%) | 1,362 (54.5%) | |
Race | ||||
White/Caucasian | 1,171 (100%) | 1,329 (100%) | 2,500 (100%) | |
Age at time of PC diagnosis | ||||
N | 1,171 | 1,329 | 2,500 | 0.3356 |
Mean (SD) | 66.1 (10.51) | 65.7 (10.71) | 65.9 (10.62) | |
Median | 67.0 | 67.0 | 67.0 | |
Q1, Q3 | 59.0, 74.0 | 58.0, 74.0 | 59.0, 74.0 | |
Range | (30.0–95.0) | (28.0–91.0) | (28.0–95.0) | |
Ever smoker | ||||
No | 627 (53.5%) | 532 (40%) | 1,159 (46.4%) | <0.0001 |
Yes | 544 (46.5%) | 797 (60%) | 1,341 (53.6%) | |
Pack years | ||||
N | 1,146 | 1,083 | 2,229 | <0.0001 |
Mean (SD) | 9.6 (17.61) | 14.8 (22.16) | 12.1 (20.12) | |
Median | 0.0 | 0.4 | 0.0 | |
Q1, Q3 | 0.0, 12.8 | 0.0, 25.0 | 0.0, 19.0 | |
Range | (0.0–125.0) | (0.0–140.0) | (0.0–140.0) | |
Diabetes > 2 years prior to study entry | ||||
No | 1,106 (94.4%) | 1,156 (87%) | 2,262 (90.5%) | <0.0001 |
Yes | 65 (5.6%) | 173 (13%) | 238 (9.5%) | |
Body mass index | ||||
N | 1,171 | 1,329 | 2,500 | <0.0001 |
Mean (SD) | 27.2 (4.63) | 28.4 (5.69) | 27.8 (5.26) | |
Median | 26.6 | 27.6 | 27.1 | |
Q1, Q3 | 24.0, 29.5 | 24.4, 31.1 | 24.3, 30.3 | |
Range | (17.7–54.2) | (15.3–59.0) | (15.3–59.0) | |
First-degree relative with pancreatic cancer | ||||
No | 1,127 (96.2%) | 1,244 (93.6%) | 2,371 (94.8%) | 0.0029 |
Yes | 44 (3.8%) | 85 (6.4%) | 129 (5.2%) |
. | Control (N = 1,171) . | Case (N = 1,329) . | Total (N = 2,500) . | P . |
---|---|---|---|---|
Sex | ||||
Female | 574 (49%) | 564 (42.4%) | 1,138 (45.5%) | 0.0010 |
Male | 597 (51%) | 765 (57.6%) | 1,362 (54.5%) | |
Race | ||||
White/Caucasian | 1,171 (100%) | 1,329 (100%) | 2,500 (100%) | |
Age at time of PC diagnosis | ||||
N | 1,171 | 1,329 | 2,500 | 0.3356 |
Mean (SD) | 66.1 (10.51) | 65.7 (10.71) | 65.9 (10.62) | |
Median | 67.0 | 67.0 | 67.0 | |
Q1, Q3 | 59.0, 74.0 | 58.0, 74.0 | 59.0, 74.0 | |
Range | (30.0–95.0) | (28.0–91.0) | (28.0–95.0) | |
Ever smoker | ||||
No | 627 (53.5%) | 532 (40%) | 1,159 (46.4%) | <0.0001 |
Yes | 544 (46.5%) | 797 (60%) | 1,341 (53.6%) | |
Pack years | ||||
N | 1,146 | 1,083 | 2,229 | <0.0001 |
Mean (SD) | 9.6 (17.61) | 14.8 (22.16) | 12.1 (20.12) | |
Median | 0.0 | 0.4 | 0.0 | |
Q1, Q3 | 0.0, 12.8 | 0.0, 25.0 | 0.0, 19.0 | |
Range | (0.0–125.0) | (0.0–140.0) | (0.0–140.0) | |
Diabetes > 2 years prior to study entry | ||||
No | 1,106 (94.4%) | 1,156 (87%) | 2,262 (90.5%) | <0.0001 |
Yes | 65 (5.6%) | 173 (13%) | 238 (9.5%) | |
Body mass index | ||||
N | 1,171 | 1,329 | 2,500 | <0.0001 |
Mean (SD) | 27.2 (4.63) | 28.4 (5.69) | 27.8 (5.26) | |
Median | 26.6 | 27.6 | 27.1 | |
Q1, Q3 | 24.0, 29.5 | 24.4, 31.1 | 24.3, 30.3 | |
Range | (17.7–54.2) | (15.3–59.0) | (15.3–59.0) | |
First-degree relative with pancreatic cancer | ||||
No | 1,127 (96.2%) | 1,244 (93.6%) | 2,371 (94.8%) | 0.0029 |
Yes | 44 (3.8%) | 85 (6.4%) | 129 (5.2%) |
aOnly subjects included in final analysis.
SNP . | Primary set . | PanScan (cohort) . | PanScan (case–control) . | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | Location . | Number cases . | Number controls . | OR . | P . | Number cases . | Number controls . | Frequency of coded allele . | OR . | P . | Number cases . | Number controls . | Frequency of coded allele . | OR . | P . |
CD101, rs10923193 | 117338325 | 1,312 | 1,169 | 0.8 | 0.000903 | 1,408 | 1,461 | 0.27 | 1.04 | 0.48 | 1,090 | 1,168 | 0.26 | 0.96 | 0.58 |
G/G | 755 (0.58) | 613 (0.52) | 1 | 1 | |||||||||||
A/G | 492 (0.42) | 460 (0.35) | 0.87 | 0.00126 | |||||||||||
A/A | 65 (0.05) | 96 (0.08) | 0.54 | 0.00126 | |||||||||||
Dominant | 0.81 | 0.013 | |||||||||||||
Recessive | 0.58 | 0.000996 | |||||||||||||
NOS1, rs3782203 | 116204794 | 1,328 | 1,171 | 1.24 | 0.0016 | 1,408 | 1,461 | 0.2 | 1 | 0.95 | 1,090 | 1,168 | 0.2 | 0.97 | 0.64 |
G/G | 780 (0.59) | 748 (0.64) | 1 | 1 | |||||||||||
A/G | 464 (0.4) | 378 (0.28) | 1.17 | 0.0733 | |||||||||||
A/A | 84 (0.06) | 45 (0.04) | 1.8 | 0.00222 | |||||||||||
Dominant | 1.23 | 0.0109 | |||||||||||||
Recessive | 1.7 | 0.005 | |||||||||||||
NOS1, rs9658350 | 116208811 | 1,327 | 1,170 | 1.24 | 0.00175 | 1,408 | 1,461 | 0.2 | 0.99 | 0.87 | 1,090 | 1,168 | 0.2 | 0.96 | 0.58 |
A/A | 774 (0.58) | 742 (0.63) | 1 | 1 | |||||||||||
G/A | 469 (0.4) | 383 (0.29) | 1.16 | 0.0779 | |||||||||||
G/G | 84 (0.06) | 45 (0.04) | 1.79 | 0.00225 | |||||||||||
Dominant | 1.23 | 0.0124 | |||||||||||||
Recessive | 1.7 | 0.005 | |||||||||||||
NOS1, rs532967 | 116216722 | 1,329 | 1,170 | 1.25 | 0.00159 | 1,408 | 1,461 | 0.18 | 0.98 | 0.76 | 1,090 | 1,168 | 0.18 | 0.92 | 0.31 |
G/G | 817 (0.61) | 780 (0.67) | 1 | 1 | |||||||||||
A/G | 447 (0.38) | 357 (0.27) | 1.19 | 0.0502 | |||||||||||
A/A | 65 (0.05) | 33 (0.03) | 1.89 | 0.00356 | |||||||||||
Dominant | 1.25 | 0.00972 | |||||||||||||
Recessive | 1.79 | 0.00733 | |||||||||||||
NOS1, rs547954 | 116238889 | 1,329 | 1,171 | 1.27 | 0.00085 | 1,408 | 1,461 | 0.17 | 0.98 | 0.79 | 1,090 | 1,168 | 0.17 | 0.90 | 0.18 |
G/G | 829 (0.62) | 796 (0.68) | 1 | 1 | |||||||||||
A/G | 437 (0.37) | 343 (0.26) | 1.21 | 0.0273 | |||||||||||
A/A | 63 (0.05) | 32 (0.03) | 1.9 | 0.00373 | |||||||||||
Dominant | 1.27 | 0.00433 | |||||||||||||
Recessive | 1.79 | 0.00882 |
SNP . | Primary set . | PanScan (cohort) . | PanScan (case–control) . | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | Location . | Number cases . | Number controls . | OR . | P . | Number cases . | Number controls . | Frequency of coded allele . | OR . | P . | Number cases . | Number controls . | Frequency of coded allele . | OR . | P . |
CD101, rs10923193 | 117338325 | 1,312 | 1,169 | 0.8 | 0.000903 | 1,408 | 1,461 | 0.27 | 1.04 | 0.48 | 1,090 | 1,168 | 0.26 | 0.96 | 0.58 |
G/G | 755 (0.58) | 613 (0.52) | 1 | 1 | |||||||||||
A/G | 492 (0.42) | 460 (0.35) | 0.87 | 0.00126 | |||||||||||
A/A | 65 (0.05) | 96 (0.08) | 0.54 | 0.00126 | |||||||||||
Dominant | 0.81 | 0.013 | |||||||||||||
Recessive | 0.58 | 0.000996 | |||||||||||||
NOS1, rs3782203 | 116204794 | 1,328 | 1,171 | 1.24 | 0.0016 | 1,408 | 1,461 | 0.2 | 1 | 0.95 | 1,090 | 1,168 | 0.2 | 0.97 | 0.64 |
G/G | 780 (0.59) | 748 (0.64) | 1 | 1 | |||||||||||
A/G | 464 (0.4) | 378 (0.28) | 1.17 | 0.0733 | |||||||||||
A/A | 84 (0.06) | 45 (0.04) | 1.8 | 0.00222 | |||||||||||
Dominant | 1.23 | 0.0109 | |||||||||||||
Recessive | 1.7 | 0.005 | |||||||||||||
NOS1, rs9658350 | 116208811 | 1,327 | 1,170 | 1.24 | 0.00175 | 1,408 | 1,461 | 0.2 | 0.99 | 0.87 | 1,090 | 1,168 | 0.2 | 0.96 | 0.58 |
A/A | 774 (0.58) | 742 (0.63) | 1 | 1 | |||||||||||
G/A | 469 (0.4) | 383 (0.29) | 1.16 | 0.0779 | |||||||||||
G/G | 84 (0.06) | 45 (0.04) | 1.79 | 0.00225 | |||||||||||
Dominant | 1.23 | 0.0124 | |||||||||||||
Recessive | 1.7 | 0.005 | |||||||||||||
NOS1, rs532967 | 116216722 | 1,329 | 1,170 | 1.25 | 0.00159 | 1,408 | 1,461 | 0.18 | 0.98 | 0.76 | 1,090 | 1,168 | 0.18 | 0.92 | 0.31 |
G/G | 817 (0.61) | 780 (0.67) | 1 | 1 | |||||||||||
A/G | 447 (0.38) | 357 (0.27) | 1.19 | 0.0502 | |||||||||||
A/A | 65 (0.05) | 33 (0.03) | 1.89 | 0.00356 | |||||||||||
Dominant | 1.25 | 0.00972 | |||||||||||||
Recessive | 1.79 | 0.00733 | |||||||||||||
NOS1, rs547954 | 116238889 | 1,329 | 1,171 | 1.27 | 0.00085 | 1,408 | 1,461 | 0.17 | 0.98 | 0.79 | 1,090 | 1,168 | 0.17 | 0.90 | 0.18 |
G/G | 829 (0.62) | 796 (0.68) | 1 | 1 | |||||||||||
A/G | 437 (0.37) | 343 (0.26) | 1.21 | 0.0273 | |||||||||||
A/A | 63 (0.05) | 32 (0.03) | 1.9 | 0.00373 | |||||||||||
Dominant | 1.27 | 0.00433 | |||||||||||||
Recessive | 1.79 | 0.00882 |
Discussion
We report the results of a case–control study evaluating the risk of inflammation-related gene variants with PC. To date, this is the largest evaluation of risk for PC that focuses primarily on genes in the inflammation pathways involving NF-κB. The 102 genes code for proinflammatory mediators, inhibitors, or activators of NF-κB. Polymorphisms of NOS1 and CD101 showed increased (NOS1) and decreased (CD101) risk association for PC. The NOS1 SNPs were in high linkage disequilibrium (LD) and located across a region of 2 LD blocks on chromosome 12. Imputed and genotyped SNPs from CD101 were located in a single LD block. Attempts to validate these data utilizing the PanScan cohort and PanScan case–control studies of PC were unsuccessful. Potential reasons for the lack of validation include the differences in study designs and accrual methods of the 3 data sets and the inability to adjust the PanScan data sets.
Conclusion
Of the 102 genes evaluated, NOS1 and CD101 may be associated with an increased and decreased risk of PC, respectively. However, these findings did not replicate in a follow-up study of 2 PC populations. Future research is needed to determine the role, if any, of NOS1 and CD101 for risk of PC.
Disclosure of Potential Conflicts of Interest
The contents of this article are solely the responsibility of the authors and do not necessarily represent the official view of NCRR or NIH. No potential conflicts of interests to disclose.
Grant Support
This publication was made possible by the Mayo Clinic SPORE in Pancreatic Cancer grant (P50 CA 102701) and grant number 1 KL2 RR024151 from the National Center for Research Resources (NCRR), a component of the NIH, and the NIH Roadmap for Medical Research. Information on NCRR is available at http://www.ncrr.nih.gov/. Information on Reengineering the Clinical Research Enterprise can be obtained from http://nihroadmap.nih.gov.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.