Abstract
BACKGROUND
We previously found that inflammation in benign prostate tissue is associated with an increased odds of prostate cancer, especially higher-grade disease. Since part of this link may be due to genetics, we evaluated the association between single nucleotide polymorphisms (SNPs) in immune response genes and prostate cancer in the placebo arm of the Prostate Cancer Prevention Trial.
METHODS
We genotyped 16 candidate SNPs in IL1β, IL2, IL4, IL6, IL8, IL10, IL12(p40), IFNG, MSR1, RNASEL, TLR4, and TNFA, and 7 tagSNPs in IL10 in 881 prostate cancer cases and 848 controls negative for cancer on an end-of-study biopsy. Cases and controls were non-Hispanic white and frequency matched on age and family history. We classified cases as lower (Gleason sum <7; N=674) and higher (7–10; N=172) grade, and used logistic regression to estimate odds ratios (OR) and 95% confidence intervals (CI) adjusting for age and family history.
RESULTS
The minor allele (C) of rs3212227 in IL12(p40) was associated with an increased risk of total (log additive: OR=1.30, 95% CI 1.10–1.53; P-trend=0.0017) and lower-grade (OR=1.36, 95% CI 1.15–1.62; P-trend=0.0004) prostate cancer. The minor allele (A) of rs4073 in IL8 was possibly associated with a decreased risk of higher-grade (OR=0.81, 95% CI 0.64–1.02; P-trend=0.07), but not total disease. None of the other candidates was associated with risk. The minor alleles of IL10 tagSNPs rs1800890 (A; OR=0.87, 95% CI: 0.75–0.99; P-trend=0.04) and rs3021094 (C; OR=1.31, 95% CI 1.03–1.66, P-trend=0.03) were associated with risk; the latter also with lower- (P-trend=0.04) and possibly higher- (P-trend=0.06) grade disease. These patterns were similar among men with PSA <2 ng/mL at biopsy.
CONCLUSION
Variation in some immune response genes may be associated with prostate cancer risk. These associations were not fully explained by PSA-associated detection bias. Our findings generally support the role of inflammation in the etiology of prostate cancer.
Keywords: Genes, Prostate Cancer, Inflammation risk
INTRODUCTION
Prostatic inflammation, especially chronic, is suspected to be a risk factor for prostate cancer [1]. Indeed, in the Prostate Cancer Prevention Trial (PCPT), we found that inflammation in benign prostate tissue, which was primarily chronic inflammation, was associated with a higher odds of prostate cancer, especially higher-grade disease [2]. In the PCPT, men were screened annually, and if not diagnosed with prostate cancer by 7-years of follow-up, underwent an “end-of-study” biopsy [3]. Thus, we were uniquely able to evaluate the association between inflammation and prostate cancer, particularly in men with a lower serum prostate specific antigen (PSA) concentration. In such men, detection bias resulting from a link between inflammation and elevated PSA, an indication for biopsy, is less likely. In the prior nested case-control study, we could not determine the timing of the presence of inflammation relative to the onset of cancer because we evaluated inflammation in the biopsy cores that were obtained to confirm or deny a diagnosis of prostate cancer.
Thus, in the PCPT, we now investigate in a prospective study – in which temporality is clear – whether germline single nucleotide polymorphisms (SNPs) in selected genes involved in inflammation and the immune response are associated with risk of prostate cancer overall and of higher- and lower-grade disease. Prior studies have investigated SNPs in genes involved in inflammatory pathways [4,5] and reported some association with prostate cancer risk. In addition, genome-wide association studies [6,7] have also identified a small number of inflammation-related SNPs as being associated with prostate cancer risk. However, none of these studies could rule out the bias that may result from the known association between some SNPs and serum PSA concentration [8–10], a commonly used screening test for prostate cancer. In our current study of 881 prostate cancer cases and 848 controls nested in the placebo arm of the trial, we genotyped 16 candidate SNPs in 12 genes – IL1β, IL2, IL4, IL6, IL8, IL10, IL12(p40), IFNG, MSR1, RNASEL, TLR4, TNFA, and 7 tagSNPs in IL10. We investigated the possibility of PSA-associated detection bias by determining serum PSA concentration as a function of genotype. Further, we evaluated the SNP-prostate cancer association in men with low serum PSA concentration; that is, men in whom detection bias resulting from the link between SNPs and elevated PSA, an indication for biopsy, is less likely.
MATERIALS AND METHODS
Study Design and Population
18,882 men enrolled in the PCPT between 1993 and1997 [3]. To be eligible, men had to be ≥55 years old and have a normal digital-rectal examination (DRE), serum PSA ≤3 ng/mL, and an American Urological Association Symptom Index <20. The participants were randomized to receive finasteride or placebo for 7 years. They completed questionnaires at the start of the trial that inquired about first-degree family history of prostate cancer, diabetes diagnosis, cigarette smoking and physical activity, and other medical and lifestyle factors. Weight and height were measured at the start of the trial from which body mass index (BMI; kg/m2) was calculated. The men were screened for prostate cancer by PSA and DRE at each of 7 annual visits. Serum remaining from the annual PSA screening and buffy coats were stored in the PCPT biorepository [11].
During the trial, a prostate biopsy was recommended when PSA was >4 ng/mL or if the DRE was abnormal. Cancers detected on these biopsies were designated as detected by “for-cause” biopsy. At the end of the trial, men who did not have a diagnosis of prostate cancer were recommended to undergo prostate biopsy, even if their PSA concentration and DRE were normal. Cancers detected on these biopsies were called “for-cause” detected if PSA >4 ng/mL or if the DRE was abnormal, but were called “end-of-study” biopsy detected if both PSA and DRE were normal. The Prostate Diagnostic Laboratory at the University of Colorado confirmed all diagnoses and determined the Gleason grade of all detected cancers.
We previously developed a case-control study nested within the PCPT, which consisted of 1,809 cases, detected either on a for-cause or end-of-study biopsy, and a sample of 1,809 men who were negative for prostate cancer on the end-of-study biopsy (controls), irrespective of whether they had a clinical indication for biopsy. Cases and controls were frequency matched on age, family history, and treatment arm, and controls were oversampled to include all non-whites. Details about the nested case-control study design have been published [11]. For the current study, we included only cases and controls who had been randomized to the placebo arm and who were non-Hispanic white. The sample size was too small to investigate SNP associations separately among other racial/ethnic groups. Thus, this study included 881 prostate cancer cases and 848 controls for whom adequate DNA was available. We classified cases as lower grade (Gleason sum 2–6; N=674) and higher grade (Gleason sum 7–10; N=172). The PCPT had 53 cases detected in the placebo arm that were Gleason 8–10 [3]. Therefore, we were unable to use a more stringent definition of higher-grade disease in this analysis. The PCPT was approved by the Institutional Review Boards at each trial site. The Johns Hopkins Bloomberg School of Public Health Institutional Review Board and the Colorado Multiple Institutional Review Board approved this inflammation study.
Genotyping
DNA was extracted from buffy coat cells at using the Qiagen M48 robot or from serum for those whom no buffy coat was available (~24% of the men). Qiagen extractions were performed at NCI Frederick, and serum DNA extractions were performed at the Wake Forest Center for Cancer Genomics using the AutoPure LS DNA Isolation Robot. We selected 16 candidate SNPs in IL1β, IL2, IL4, IL6, IL8, IL10, IL12(p40), IFNG, MSR1, RNASEL, TLR4, and TNFA based on their role in the immune response. In general, we favored genes involved in innate immunity and T cell activation/response and then prioritized the SNPs within those genes based on evidence that they influence production or activity of the gene product, their association with prostate cancer, and their association with gastric cancer and colitis. We also selected and genotyped 7 tagSNPs in IL10 as previously described [12]. Details on known or expected implications of the minor allele of each SNP on the gene product or disease risk are shown in Supplement Table 1. Genotyping was performed using the Illumina VeraCode GoldenGate 384-plex platform (at the University of Texas Health Science Center at San Antonio) or using the Sequenom MassARRAY platform (at Wake Forest School of Medicine). To be included in the analysis, a SNP had to have <5.5% missing. Genotyping failed for rs2243250 in IL4, rs3024498 in IL10, and rs2430561 in IFNG for DNA extracted from serum, and rs1800871 in IL10 was not run for DNA from serum; data for DNA extracted from buffy coat are included in the analysis. The ability to quantify SNP data was slightly higher for DNA extracted buffy coats than from serum (data not shown). We distributed duplicate samples equal to 5% of the total sample throughout the plates assess genotyping reliability. Concordance was ≥93% (mean 97%) for buffy coat DNA and 100% for serum DNA, after excluding non-evaluable samples. Three of the SNPs were not in Hardy-Weinberg equilibrium (HWE) – rs3212227 in IL12(p40) (p=0.001), rs1800872 in IL10 (p=0.037), and rs4986790 in TLR4 (p=0.014). We confirmed that the deviations in the genotype frequencies from the expected were not substantial, and thus we included these SNPs in the analysis.
Statistical Analysis
We compared baseline demographic and lifestyle characteristics of prostate cancer cases and controls by t-tests for continuous variables and χ2 tests for categorical variables. Next, we used logistic regression to estimate odds ratios (OR) of total, lower, and higher-grade prostate cancer and 95% confidence intervals (CI) adjusting for the frequency-matching factors age and family history. Then, we considered the possibility of detection bias resulting from any link between the SNPs and serum PSA, first, by calculating mean serum PSA concentration by genotype in the controls and testing for trend across genotype using linear regression. Then, we reduced the possibility of PSA-associated detection bias by restricting the analysis to cases and controls with a serum PSA concentration <2 ng/mL (a concentration substantially lower than the conventional cutpoint to prompt biopsy) within one year prior to biopsy and estimating the SNP-prostate cancer association using logistic regression. Statistical analyses were performed using the R packages SNPassoc (http://www.creal.cat/jrgonzalez/software.htm) and haplo.stats (http://cran.r-project.org/web/packages/haplo.stats/index.html), and SAS (version 9.2, Cary, NC) by the SWOG Statistical Center at the Fred Hutchinson Cancer Research Center, Seattle, WA. Two-sided tests were performed, and p-values <0.05 were considered to be statistically significant.
RESULTS
Table I. shows characteristics of the prostate cases and controls. Compared with controls, cases were less likely to report a diagnosis of diabetes (P=0.01), more likely to have a normal body mass index (P=0.05), and had higher median baseline PSA (1.4 ng/mL in cases versus 1.0 ng/mL in controls, p<0.0001), but did not differ on physical activity, smoking status, or on the frequency-matching factors age and family history of prostate cancer. The median age of both cases and controls was 63 years old. Of the cases, 42.0% were detected on a “for-cause” biopsy, 24.9% had a PSA level ≥2 ng/mL at baseline, and 20.3% had higher-grade disease.
Table I.
Characteristics* of prostate cancer and controls, placebo arm of the PCPT
| Cases | Controls | P-value | |
|---|---|---|---|
| Number of Men | 881 | 848 | |
| Age at Baseline Category (%)* | |||
| 55–59 years | 236 (26.8) | 212 (25.0) | 0.83 |
| 60–64 years | 288 (32.7) | 277 (32.7) | |
| 65–69 years | 208 (23.6) | 208 (24.5) | |
| 70+ years | 149 (16.9) | 151 (17.8) | |
| Family History of Prostate Cancer (%)** | 181 (20.5) | 182 (21.5) | 0.64 |
| Baseline Serum PSA Category (%) | |||
| 0.0–1.0 ng/mL | 285 (32.3) | 437 (51.5) | <0.001 |
| 1.1–2.0 ng/mL | 377 (42.8) | 293 (34.6) | |
| 2.1–3.0 ng/mL | 219 (24.9) | 118 (13.9) | |
| History of Diabetes (%) | 30 (3.4) | 53 (6.3) | 0.01 |
| Physical Activity Category (%) | |||
| Sedentary | 134 (15.3) | 132 (15.6) | 0.33 |
| Light | 355 (40.4) | 355 (42.0) | |
| Moderate | 308 (35.1) | 265 (31.4) | |
| Active | 81 (9.2) | 93 (11.0) | |
| BMI Category (%) | |||
| Normal (<25 kg/m2) | 263 (30.1) | 209 (25.0) | 0.05 |
| Overweight (25 to <30 kg/m2) | 454 (52.0) | 457 (54.6) | |
| Obese (≥30 kg/m2) | 156 (17.9) | 171 (20.4) | |
| Smoking Status (%) | |||
| Never Smoker | 323 (36.7) | 299 (35.3) | 0.77 |
| Current Smoke | 52 (5.9) | 55 (6.5) | |
| Former Smoker | 506 (57.4) | 494 (58.3) | |
| Cancer Detected on a For-Cause Biopsy (%) | 370 (42.0) | - | - |
| Gleason Sum Category (%) | |||
| Lower grade, 2–6 | 674 (79.7) | - | - |
| Higher grade, 7–10 | 172 (20.3) | - | - |
Restricted to whites
Frequency matching variable
Candidate SNPs and prostate cancer risk
Men with two copies of the minor allele (C) of rs3212227 in IL12(p40) had a higher risk of prostate cancer (OR=1.73, 95% CI 1.11–2.67; P-trend=0.0017; Table II), especially lower-grade disease (OR=1.87, 95% CI 1.18–2.95; P-trend=0.0004; Table III), than men with one or two copies the major allele (A). The minor allele (A) of rs4073 in IL8 was possibly associated with a lower risk of lower-grade disease (log-additive OR=0.81, 95% CI 0.64–1.02; P-trend=0.07; Table IV), but not prostate cancer overall (Table II). Carrying two copies of the minor allele (A) of rs486907 in RNASEL was possibly associated with a lower risk of prostate cancer, especially higher-grade disease (OR=0.51, 95% CI 0.28–0.94; P-trend=0.12; Table III); no association was present for carrying one copy of the minor allele. None of the other candidates, including in IL10, was significantly associated with prostate cancer risk or grade of disease (Table II and Table III).
Table II.
Association between SNPs in genes involved in the immune response and prostate cancer risk, placebo arm of the PCPT
| Number of Minor Alleles
|
Log Additive | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gene | dbSNP | None | 1 copy | 2 Copies | ||||||||
|
|
||||||||||||
| Cases/Controls | Genotype | OR | Cases/Controls | Genotype | OR (95% CI) | Cases/Controls | Genotype | OR (95%CI) | OR (95% CI) | P-trend | ||
| IL1β | rs1143634 | 517/504 | C/C | 1 (ref) | 307/278 | C/T | 1.07 (0.88–1.32) | 30/42 | T/T | 0.69 (0.43–1.13) | 0.97 (0.82–1.14) | 0.7 |
| IL1β | rs1143627 | 383/371 | T/T | 1 (ref) | 349/347 | C/T | 0.98 (0.79–1.20) | 105/83 | C/C | 1.23 (0.89–1.69) | 1.06 (0.92–1.23) | 0.4 |
| IL2 | rs2069762 | 429/413 | T/T | 1 (ref) | 357/341 | G/T | 1.01 (0.83–1.23) | 71/71 | G/G | 0.96 (0.68–1.38) | 0.99 (0.86–1.15) | 0.9 |
| IL4 | rs2243250 | 467/452 | C/C | 1 (ref) | 160/192 | C/T | 0.80 (0.63–1.03) | 12/10 | T/T | 1.16 (0.50–2.71) | 0.86 (0.69–1.07) | 0.2 |
| IL6 | rs1800795 | 287/287 | G/G | 1 (ref) | 404/406 | C/G | 1.00 (0.80–1.23) | 166/141 | C/C | 1.18 (0.89–1.56) | 1.07 (0.94–1.23) | 0.3 |
| IL6 | rs1800797 | 302/298 | G/G | 1 (ref) | 406/396 | A/G | 1.01 (0.82–1.25) | 153/134 | A/A | 1.13 (0.85–1.50) | 1.05 (0.92–1.21) | 0.5 |
| IL8 | rs4073 | 260/249 | T/T | 1 (ref) | 418/404 | A/T | 0.99 (0.79–1.23) | 177/186 | A/A | 0.91 (0.70–1.19) | 0.96 (0.84–1.09) | 0.5 |
| IL10 | rs1800871 | 381/408 | C/C | 1 (ref) | 216/217 | C/T | 1.07 (0.85–1.35) | 35/34 | T/T | 1.10 (0.67–1.81) | 1.06 (0.88–1.27) | 0.5 |
| IL10 | rs1800872 | 514/511 | C/C | 1 (ref) | 292/269 | A/C | 1.08 (0.88–1.33) | 55/47 | A/A | 1.16 (0.77–1.75) | 1.08 (0.92–1.26) | 0.3 |
| IL10 | rs1800896 | 218/204 | A/A | 1 (ref) | 449/429 | A/G | 0.98 (0.78–1.24) | 200/203 | G/G | 0.92 (0.70–1.21) | 0.96 (0.84–1.10) | 0.6 |
| IL10 | rs3024498* | 338/362 | A/A | 1 (ref) | 253/254 | A/G | 1.07 (0.85–1.34) | 48/50 | G/G | 1.02 (0.67–1.56) | 1.04 (0.87–1.23) | 0.7 |
| IL10 | rs3024496* | 223/208 | T/T | 1 (ref) | 439/417 | C/T | 0.98 (0.78–1.24) | 201/05 | C/C | 0.91 (0.70–1.20) | 0.96 (0.83–1.09) | 0.5 |
| IL10 | rs3024509* | 765/741 | T/T | 1 (ref) | 93/87 | C/T | 1.03 (0.76–1.41) | 5/4 | C/C | 1.20(0.32–4.48) | 1.04 (0.79–1.39) | 0.8 |
| IL10 | rs1554286* | 585/588 | C/C | 1 (ref) | 258/226 | C/T | 1.15 (0.93–1.42) | 31/31 | T/T | 1.00(0.60–1.67) | 1.09(0.92–1.29) | 0.3 |
| IL10 | rs3021094* | 713/716 | A/A | 1 (ref) | 152/124 | A/C | 1.23 (0.95–1.60) | 10/3 | C/C | 3.40 (0.93–12.4) | 1.31 (1.03–1.66) | 0.03 |
| IL10 | rs1800894* | 809/793 | G/G | 1 (ref) | 60/49 | A/G | 1.20 (0.81–1.77) | 2/0 | A/A | - | 1.27 (0.87–1.85) | 0.2 |
| IL10 | rs1800890* | 321/274 | T/T | 1 (ref) | 417/419 | A/T | 0.85 (0.69–1.05) | 131/148 | A/A | 0.75 (0.57–1.00) | 0.87 (0.75–0.99) | 0.04 |
| IL12(p40) | rs3212227 | 535/568 | A/A | 1 (ref) | 274/227 | A/C | 1.28 (1.04–1.59) | 57/35 | C/C | 1.73 (1.11–2.67) | 1.30 (1.10–1.53) | 0.0017 |
| IFNG | rs2430561 | 174/190 | T/T | 1 (ref) | 326/326 | A/T | 1.09 (0.84–1.41) | 135/136 | A/A | 1.08 (0.79–1.48) | 1.04 (0.89–1.22) | 0.6 |
| MSR1 | rs3747531 | 779/734 | G/G | 1 (ref) | 93/99 | C/G | 0.89 (0.66–1.20) | 2/4 | C/C | 0.47 (0.09–2.58) | 0.86 (0.65–1.14) | 0.3 |
| RNASEL | rs486907 | 352/330 | G/G | 1 (ref) | 407/372 | A/G | 1.03 (0.84–1.26) | 105/129 | A/A | 0.76 (0.57–1.03) | 0.91 (0.79–1.05) | 0.2 |
| TLR4 | rs4986790 | 768/741 | A/A | 1 (ref) | 94/82 | A/G | 1.11 (0.81–1.52) | 5/7 | G/G | 0.69 (0.22–2.19) | 1.04 (0.79–1.37) | 0.8 |
| TNFA | rs1800629 | 607/580 | G/G | 1 (ref) | 238/235 | A/G | 0.97 (0.78–1.20) | 21/23 | A/A | 0.88 (0.48–1.60) | 0.96 (0.80–1.15) | 0.6 |
TagSNP
Table III.
Association between SNPs in genes involved in immune response and risk of lower- and higher-grade prostate cancer, placebo arm of the PCPT
| Number of Minor Alleles
|
Log Additive | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gene | dbSNP | None | 1 copy | 2 Copies | ||||||||
|
|
||||||||||||
| Cases/Controls | Genotype | OR | Cases/Controls | Genotype | OR (95% CI) | Cases/Controls | Genotype | OR (95%CI) | OR (95% CI) | P-trend | ||
|
Lower grade (Gleason sum 2–6)
| ||||||||||||
| IL1β | rs1143634 | 396/504 | C/C | 1 (ref) | 236/278 | C/T | 1.07 (0.86–1.33) | 20/42 | T/T | 0.60 (0.35–1.04) | 0.95 (0.79–1.13) | 0.5 |
| IL1β | rs1143627 | 288/371 | T/T | 1 (ref) | 281/347 | C/T | 1.04 (0.84–1.30) | 73/83 | C/C | 1.13 (0.80–1.61) | 1.06 (0.90–1.24) | 0.5 |
| IL2 | rs2069762 | 334/413 | T/T | 1 (ref) | 260/341 | G/T | 0.95 (0.76–1.18) | 59/71 | G/G | 1.03 (0.71–1.49) | 0.99 (0.84–1.16) | 0.9 |
| IL4 | rs2243250 | 362/452 | C/C | 1 (ref) | 128/192 | C/T | 0.83 (0.64–1.08) | 8/10 | T/T | 0.99 (0.39–2.55) | 0.86 (0.68–1.09) | 0.2 |
| IL6 | rs1800795 | 215/287 | G/G | 1 (ref) | 314/406 | C/G | 1.03 (0.82–1.30) | 126/141 | C/C | 1.20 (0.89–1.61) | 1.08 (0.94–1.25) | 0.3 |
| IL6 | rs1800797 | 229/298 | G/G | 1 (ref) | 314/396 | A/G | 1.03 (0.82–1.30) | 118/134 | A/A | 1.15 (0.85–1.56) | 1.07 (0.92–1.23) | 0.4 |
| IL8 | rs4073 | 182/249 | T/T | 1 (ref) | 326/404 | A/T | 1.10 (0.86–1.40) | 145/186 | A/A | 1.06 (0.80–1.42) | 1.04 (0.90–1.20) | 0.6 |
| IL10 | rs1800871 | 292/408 | C/C | 1 (ref) | 171/217 | C/T | 1.11 (0.86–1.42) | 28/34 | T/T | 1.15 (0.68–1.93) | 1.09 (0.90–1.33) | 0.4 |
| IL10 | rs1800872 | 390/511 | C/C | 1 (ref) | 226/269 | A/C | 1.11 (0.89–1.38) | 43/47 | A/A | 1.20 (0.78–1.86) | 1.10 (0.93–1.30) | 0.3 |
| IL10 | rs1800896 | 165/204 | A/A | 1 (ref) | 344/429 | A/G | 0.99 (0.77–1.28) | 153/203 | G/G | 0.93 (0.69–1.25) | 0.96 (0.83–1.12) | 0.6 |
| IL10 | rs3024498* | 264/362 | A/A | 1 (ref) | 195/254 | A/G | 1.05 (0.82–1.34) | 37/50 | G/G | 1.01 (0.64–1.58) | 1.02 (0.85–1.23) | 0.8 |
| IL10 | rs3024496* | 167/208 | T/T | 1 (ref) | 339/417 | C/T | 1.01 (0.79–1.30) | 155/205 | C/C | 0.94 (0.70–1.25) | 0.97 (0.84–1.12) | 0.7 |
| IL10 | rs3024509* | 583/741 | T/T | 1 (ref) | 73/87 | C/T | 1.06 (0.76–1.48) | 3/4 | C/C | 0.94 (0.21–4.21) | 1.05 (0.77–1.42) | 0.8 |
| IL10 | rs1554286* | 443/588 | C/C | 1 (ref) | 202/226 | C/T | 1.19 (0.95–1.50) | 24/31 | T/T | 1.02 (0.59–1.77) | 1.12 (0.93–1.34) | 0.2 |
| IL10 | rs3021094* | 545/716 | A/A | 1 (ref) | 117/124 | A/C | 1.24 (0.94–1.64) | 7/3 | C/C | 3.25 (0.84–12.68) | 1.31 (1.02–1.70) | 0.04 |
| IL10 | rs1800894* | 614/793 | G/G | 1 (ref) | 49/49 | A/G | 1.28 (0.85–1.93) | 2/0 | A/A | - | 1.37 (0.93–2.04) | 0.11 |
| IL10 | rs1800890* | 239/274 | T/T | 1 (ref) | 326/419 | A/T | 0.89 (0.71–1.12) | 101/148 | A/A | 0.78 (0.57–1.06) | 0.88 (0.76–1.03) | 0.10 |
| IL12(p40) | rs3212227 | 400/568 | A/A | 1 (ref) | 217/227 | A/C | 1.36 (1.09–1.71) | 46/35 | C/C | 1.87 (1.18–2.95) | 1.36 (1.15–1.62) | 0.0004 |
| IFNG | rs2430561 | 134/190 | T/T | 1 (ref) | 254/326 | A/T | 1.10 (0.83–1.44) | 107/136 | A/A | 1.11 (0.79–1.56) | 1.06 (0.89–1.25) | 0.5 |
| MSR1 | rs3747531 | 597/734 | G/G | 1 (ref) | 72/99 | C/G | 0.90 (0.65–1.24) | 0/4 | C/C | - | 0.83 (0.61–1.13) | 0.2 |
| RNASEL | rs486907 | 268/330 | G/G | 1 (ref) | 307/372 | A/G | 1.02 (0.81–1.27) | 86/129 | A/A | 0.82 (0.60–1.12) | 0.93 (0.80–1.08) | 0.3 |
| TLR4 | rs4986790 | 593/741 | A/A | 1 (ref) | 70/82 | A/G | 1.07 (0.76–1.50) | 4/7 | G/G | 0.73 (0.21–2.49) | 1.01 (0.75–1.37) | 0.9 |
| TNFA | rs1800629 | 470/580 | G/G | 1 (ref) | 179/235 | A/G | 0.95 (0.75–1.19) | 14/23 | A/A | 0.76 (0.39–1.49) | 0.92 (0.76–1.13) | 0.4 |
|
| ||||||||||||
|
Higher grade (Gleason sum 7–10)
| ||||||||||||
| IL1β | rs1143634 | 101/504 | C/C | 1 (ref) | 59/278 | C/T | 1.08 (0.76–1.53) | 8/42 | T/T | 0.96 (0.44–2.12) | 1.03 (0.78–1.37) | 0.8 |
| IL1β | rs1143627 | 79/371 | T/T | 1 (ref) | 53/347 | C/T | 0.71 (0.49–1.03) | 30/83 | C/C | 1.70 (1.05–2.75) | 1.13 (0.88–1.44) | 0.3 |
| IL2 | rs2069762 | 82/413 | T/T | 1 (ref) | 78/341 | G/T | 1.16 (0.82–1.63) | 10/71 | G/G | 0.71 (0.35–1.44) | 0.98 (0.76–1.27) | 0.9 |
| IL4 | rs2243250 | 91/452 | C/C | 1 (ref) | 26/192 | C/T | 0.68 (0.43–1.09) | 3/10 | T/T | 1.50 (0.40–5.57) | 0.80 (0.53–1.20) | 0.3 |
| IL6 | rs1800795 | 53/287 | G/G | 1 (ref) | 78/406 | C/G | 1.04 (0.71–1.53) | 36/141 | C/C | 1.39 (0.87–2.22) | 1.16 (0.92–1.47) | 0.2 |
| IL6 | rs1800797 | 54/298 | G/G | 1 (ref) | 81/396 | A/G | 1.13 (0.77–1.64) | 30/134 | A/A | 1.23 (0.75–2.01) | 1.11 (0.88–1.41) | 0.4 |
| IL8 | rs4073 | 61/249 | T/T | 1 (ref) | 78/404 | A/T | 0.79 (0.55–1.15) | 30/186 | A/A | 0.65 (0.41–1.05) | 0.81 (0.64–1.02) | 0.07 |
| IL10 | rs1800871 | 80/408 | C/C | 1 (ref) | 35/217 | C/T | 0.82 (0.53–1.25) | 5/34 | T/T | 0.78 (0.30–2.07) | 0.84 (0.60–1.19) | 0.3 |
| IL10 | rs1800872 | 105/511 | C/C | 1 (ref) | 54/269 | A/C | 0.96 (0.67–1.38) | 8/47 | A/A | 0.86 (0.39–1.88) | 0.95 (0.71–1.26) | 0.7 |
| IL10 | rs1800896 | 41/204 | A/A | 1 (ref) | 90/429 | A/G | 1.05 (0.70–1.57) | 39/203 | G/G | 0.97 (0.60–1.56) | 0.98 (0.78–1.25) | 0.9 |
| IL10 | rs3024498* | 61/362 | A/A | 1 (ref) | 51/254 | A/G | 1.18 (0.79–1.77) | 10/50 | G/G | 1.21 (0.58–2.52) | 1.13 (0.84–1.53) | 0.4 |
| IL10 | rs3024496* | 44/208 | T/T | 1 (ref) | 85/417 | C/T | 0.97 (0.65–1.45) | 38/205 | C/C | 0.89 (0.55–1.43) | 0.94 (0.74–1.19) | 0.6 |
| IL10 | rs3024509* | 153/741 | T/T | 1 (ref) | 16/87 | C/T | 0.89 (0.51–1.57) | 1/4 | C/C | 1.29 (0.14–11.71) | 0.93 (0.56–1.56) | 0.8 |
| IL10 | rs1554286* | 119/588 | C/C | 1 (ref) | 46/226 | C/T | 1.00 (0.69–1.45) | 5/31 | T/T | 0.82 (0.31–2.16) | 0.96 (0.71–1.31) | 0.8 |
| IL10 | rs3021094* | 137/716 | A/A | 1 (ref) | 31/124 | A/C | 1.31 (0.85–2.02) | 3/3 | C/C | 5.21 (1.04–26.12) | 1.46 (0.99–2.16) | 0.06 |
| IL10 | rs1800894* | 162/793 | G/G | 1 (ref) | 9/49 | A/G | 0.90 (0.43–1.86) | 0/0 | A/A | - | 0.90 (0.43–1.86) | 0.8 |
| IL10 | rs1800890* | 65/274 | T/T | 1 (ref) | 78/419 | A/T | 0.79 (0.55–1.13) | 25/148 | A/A | 0.72 (0.43–1.19) | 0.83 (0.65–1.06) | 0.14 |
| IL12(p40) | rs3212227 | 110/568 | A/A | 1 (ref) | 48/227 | A/C | 1.08 (0.75–1.57) | 10/35 | C/C | 1.49 (0.72–3.10) | 1.15 (0.87–1.53) | 0.3 |
| IFNG | rs2430561 | 33/190 | T/T | 1 (ref) | 63/326 | A/T | 1.12 (0.71–1.78) | 24/136 | A/A | 1.02 (0.58–1.81) | 1.02 (0.77–1.35) | 0.9 |
| MSR1 | rs3747531 | 151/734 | G/G | 1 (ref) | 19/99 | C/G | 0.93 (0.55–1.56) | 1/4 | C/C | 1.20 (0.13–10.84) | 0.95 (0.59–1.54) | 0.8 |
| RNASEL | rs486907 | 70/330 | G/G | 1 (ref) | 84/372 | A/G | 1.06 (0.75–1.51) | 14/129 | A/A | 0.51 (0.28–0.94) | 0.82 (0.64–1.05) | 0.12 |
| TLR4 | rs4986790 | 144/741 | A/A | 1 (ref) | 21/82 | A/G | 1.32 (0.79–2.20) | 1/7 | G/G | 0.72 (0.09–5.89) | 1.19 (0.76–1.87) | 0.5 |
| TNFA | rs1800629 | 112/580 | G/G | 1 (ref) | 49/235 | A/G | 1.07 (0.74–1.54) | 7/23 | A/A | 1.54 (0.64–3.68) | 1.13 (0.84–1.53) | 0.4 |
Table IV.
Association between IL10 haplotypes and risk of total, lower-, and higher-grade prostate cancer, placebo arm of the PCPT
| Prostate cancer | tagSNP in IL10 | Estimated haplotype frequencies | OR (95% CI) | P | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||||
| rs3024498 | rs3024496 | rs3024509 | rs1554286 | rs3021094 | rs1800894 | rs1800890 | Case | Control | |||
| Total | |||||||||||
| A | T | T | C | A | G | T | 0.30 | 0.30 | 1 (ref) | ||
| A | C | C | C | A | A | T | 0.04 | 0.03 | 1.25 (0.84–1.86) | 0.3 | |
| A | C | C | C | A | G | T | 0.02 | 0.03 | 0.85 (0.55–1.31) | 0.5 | |
| A | C | T | C | A | G | A | 0.14 | 0.18 | 0.81 (0.65–1.02) | 0.07 | |
| A | T | T | C | C | G | T | 0.02 | 0.02 | 1.17 (0.70–1.95) | 0.5 | |
| A | T | T | T | A | G | T | 0.11 | 0.11 | 0.96 (0.75–1.21) | 0.7 | |
| A | T | T | T | C | G | T | 0.07 | 0.06 | 1.29 (0.96–1.73) | 0.095 | |
| G | C | T | C | A | G | A | 0.24 | 0.24 | 0.98 (0.81–1.19) | 0.8 | |
| G | C | T | C | A | G | T | 0.04 | 0.03 | 1.39 (0.91–2.10) | 0.13 | |
| All rare haplotypes combined | 1.04 (0.40–2.69) | 0.9 | |||||||||
|
| |||||||||||
| Score test | 0.44 | ||||||||||
|
| |||||||||||
| Lower grade (Gleason sum 2–6) | |||||||||||
| A | T | T | C | A | G | T | 0.30 | 0.30 | 1 (ref) | ||
| A | C | C | C | A | A | T | 0.04 | 0.03 | 1.36 (0.89–2.06) | 0.2 | |
| A | C | C | C | A | G | T | 0.02 | 0.03 | 0.77 (0.47–1.25) | 0.3 | |
| A | C | T | C | A | G | A | 0.15 | 0.18 | 0.85 (0.67–1.08) | 0.2 | |
| A | T | T | C | C | G | T | 0.02 | 0.02 | 1.07 (0.61–1.87) | 0.8 | |
| A | T | T | T | A | G | T | 0.11 | 0.11 | 0.98 (0.76–1.26) | 0.9 | |
| A | T | T | T | C | G | T | 0.08 | 0.06 | 1.33 (0.97–1.82) | 0.074 | |
| G | C | T | C | A | G | A | 0.24 | 0.24 | 0.99 (0.81–1.22) | 0.9 | |
| G | C | T | T | A | G | T | #N/A | 0.00 | 1.26 (0.79–2.01) | 0.33 | |
| All rare haplotypes combined | 0.91 (0.25–3.33) | 0.9 | |||||||||
|
| |||||||||||
| Score test | 0.69 | ||||||||||
|
| |||||||||||
| Higher-grade (Gleason sum 7–10) | |||||||||||
| A | T | T | C | A | G | T | 0.31 | 0.30 | 1 (ref) | ||
| A | C | C | C | A | A | T | 0.03 | 0.03 | 0.88 (0.41–1.87) | 0.7 | |
| A | C | C | C | A | G | T | 0.03 | 0.03 | 0.93 (0.45–1.93) | 0.8 | |
| A | C | T | C | A | G | A | 0.12 | 0.18 | 0.71 (0.46–1.09) | 0.11 | |
| A | T | T | C | C | G | T | 0.03 | 0.02 | 1.70 (0.79–3.68) | 0.2 | |
| A | T | T | T | A | G | T | 0.09 | 0.11 | 0.77 (0.49–1.21) | 0.3 | |
| A | T | T | T | C | G | T | 0.07 | 0.06 | 1.30 (0.79–2.13) | 0.3 | |
| G | C | T | C | A | G | A | 0.25 | 0.24 | 0.98 (0.70–1.38) | 0.9 | |
| G | C | T | C | A | G | T | 0.06 | 0.03 | 2.07 (1.13–3.81) | 0.019 | |
| All rare haplotypes combined | 1.85 (0.53–6.50) | 0.3 | |||||||||
|
| |||||||||||
| Score test | 0.07 | ||||||||||
IL10 tagSNPs and prostate cancer risk
Two of seven tagSNPs in IL10 were statistically significantly associated with prostate cancer (rs3021094 and rs1800890) (Table II). The log-additive associations were OR=1.31 (95% CI 1.03–1.66; P-trend=0.03) for the minor allele (C) of rs3021094 and OR=0.87 (95% CI 0.75–0.99; P-trend=0.04) for the minor allele (A) of rs1800890. These associations were similar to overall for lower- and higher-grade disease (Table III).
IL10 tagSNPs were used to construct haplotypes. Compared with the most common, haplotype, ACTCAGA was possibly inversely associated with total (OR=0.81, 95% CI 0.65–1.02), lower- and higher-grade prostate cancer (Table IV). Haplotype GCTCAGT was positively associated with higher-grade disease (OR=2.07, 95% CI 1.13–3.81), but not with prostate cancer overall or lower-grade disease. Haplotype ATTTCGT was possibly positively associated with each outcome. While the haplotype distribution did not differ between total cases and controls (score test P=0.44) or lower-grade cases and controls (score test P=0.69), it possibly differed between higher-grade cases and controls (P=0.07; Table IV),
Serum PSA concentration by genotype in controls
Mean serum PSA concentration within one year prior to biopsy by genotype in controls is presented in Table V. PSA concentration increased with increasing number of minor alleles of rs2069762 in IL2 (P-trend=0.03). Mean serum PSA concentration was possibly lower with one or two copies of the minor allele of rs1800795 (P-trend=0.06) and rs1800797 (P-trend=0.09) in IL6. PSA concentration differed by genotype for two of the IL10 tagSNPs: men with at least one copy of the minor allele of rs3021094 (P-trend=0.05) and heterozygous for rs1800894 (p=0.002; none of the controls carried two copies of the minor allele) had higher PSA concentration compared with men with two copies of the major allele.
Table V.
Mean serum PSA concentration at the end of study biopsy across genotype for genes involved in the immune response, controls in the placebo arm of the PCPT
| Gene | SNP | Genotype | N | Mean serum PSA (ng/mL) | P-trend |
|---|---|---|---|---|---|
| IL1β | rs1143634 | C/C | 496 | 1.94 | 0.86 |
| C/T | 274 | 2.03 | |||
| T/T | 41 | 2.01 | |||
|
| |||||
| IL1β | rs1143627 | T/T | 364 | 2.02 | 0.56 |
| T/C | 344 | 2.05 | |||
| C/C | 80 | 1.42 | |||
|
| |||||
| IL2 | rs2069762 | T/T | 408 | 1.72 | 0.03 |
| T/G | 336 | 1.90 | |||
| G/G | 68 | 3.86 | |||
|
| |||||
| IL4 | rs2243250 | C/C | 445 | 2.17 | 0.78 |
| C/T | 191 | 1.79 | |||
| T/T | 10 | 3.71 | |||
|
| |||||
| IL6 | rs1800795 | G/G | 280 | 2.54 | 0.06** |
| G/C | 401 | 1.68 | |||
| C/C | 140 | 1.62 | |||
|
| |||||
| IL6 | rs1800797 | G/G | 291 | 2.52 | 0.09** |
| G/A | 391 | 1.66 | |||
| A/A | 133 | 1.72 | |||
|
| |||||
| IL8 | rs4073 | T/T | 244 | 2.01 | 0.64 |
| T/A | 398 | 2.01 | |||
| A/A | 184 | 1.74 | |||
|
| |||||
| IL10 | rs1800871 | C/C | 403 | 1.87 | 0.48 |
| C/T | 215 | 2.48 | |||
| T/T | 33 | 1.76 | |||
|
| |||||
| IL10 | rs1800872 | C/C | 504 | 1.79 | 0.42 |
| C/A | 266 | 2.35 | |||
| A/A | 44 | 1.70 | |||
|
| |||||
| IL10 | rs1800896 | A/A | 199 | 1.88 | 0.80 |
| A/G | 423 | 1.97 | |||
| G/G | 201 | 2.03 | |||
|
| |||||
| IL10 | rs3024496* | T/T | 203 | 1.77 | 0.57 |
| T/C | 411 | 1.99 | |||
| C/C | 204 | 2.09 | |||
|
| |||||
| IL10 | rs1800894* | G/G | 782 | 1.81 | 0.002 |
| G/A | 47 | 4.40 | |||
| A/A | 0 | - | |||
|
| |||||
| IL10 | rs1800890* | T/T | 268 | 2.21 | 0.89 |
| T/A | 412 | 1.67 | |||
| A/A | 148 | 2.31 | |||
|
| |||||
| IL10 | rs3024509* | T/T | 731 | 1.83 | 0.10 |
| T/C | 86 | 3.01 | |||
| C/C | 3 | 1.43 | |||
|
| |||||
| IL10 | rs1554286* | C/C | 581 | 1.86 | 0.66 |
| C/T | 222 | 2.19 | |||
| T/T | 29 | 1.68 | |||
|
| |||||
| IL10 | rs3021094* | A/A | 708 | 1.79 | 0.05 |
| A/C | 120 | 2.88 | |||
| C/C | 2 | 2.85 | |||
|
| |||||
| IL10 | rs3024498* | A/A | 357 | 2.16 | 0.76 |
| A/G | 251 | 1.71 | |||
| G/G | 50 | 3.30 | |||
|
| |||||
| IL12(p40) | rs3212227 | A/A | 558 | 2.03 | 0.95 |
| A/C | 226 | 1.63 | |||
| C/C | 33 | 3.21 | |||
|
| |||||
| IFNG | rs2430561 | T/T | 189 | 1.52 | 0.30 |
| T/A | 320 | 2.10 | |||
| A/A | 135 | 1.88 | |||
|
| |||||
| MSR1 | rs3747531 | G/G | 721 | 1.99 | 0.71 |
| G/C | 99 | 1.77 | |||
| C/C | 4 | 1.70 | |||
|
| |||||
| RNASEL | rs486907 | G/G | 325 | 2.06 | 0.62 |
| G/A | 367 | 1.93 | |||
| A/A | 126 | 1.77 | |||
|
| |||||
| TLR4 | rs4986790 | A/A | 730 | 2.03 | 0.40 |
| A/G | 81 | 1.54 | |||
| G/G | 7 | 1.04 | |||
|
| |||||
| TNFA | rs1800629 | G/G | 571 | 2.10 | 0.33 |
| G/A | 232 | 1.65 | |||
| A/A | 22 | 1.69 | |||
TagSNP
In a post-hoc analysis combining genotypes with similar concentrations, P=0.04 for both rs1800795 and rs1800797.
SNPs and prostate cancer risk in men with PSA <2 ng/mL
In a subanalysis restricted to cases and controls with PSA <2 ng/mL within one year of prior to biopsy, the minor allele of rs3212227 in IL12(p40) remained positively associated with total and lower-grade disease (Supplement Table 2). The minor allele of rs4073 in IL8 was even more inversely associated with a higher-grade disease in men with low PSA than in all men; but was still not associated with prostate cancer overall and lower-grade disease. For rs486907 in RNASEL, when restricted to men with low PSA, the association for carrying two copies of the minor allele was still in the inverse direction, including for higher-grade disease, but was not statistically significant (Supplement Table 3). Unlike in the primary analysis, two of the candidates (rs1800871, rs1800872) in IL10 were positively associated with total and lower-grade disease in men with low PSA. Also, the associations of tagSNP rs3021094 in IL10 with total, lower- and higher-grade disease were similar in men with low PSA and all men. For tagSNP rs1800890 in IL10, the magnitude of the associations for total and lower-grade disease were similar in men with low PSA and all men, but the OR was null for higher-grade disease. In contrast to all men, two additional tagSNPs in IL10 were possibly associated with total and lower-grade disease (rs1554286) or lower grade disease (rs3024496) in men with low PSA. While the distribution of the haplotypes did not differ between the prostate cancer cases and controls or the lower- or higher-grade cases and controls, the same haplotypes that were associated with prostate cancer overall or with lower- or higher-grade disease persisted in the men with low PSA, and in some haplotypes were more strongly associated with risk (Supplement Table 3). When restricting to men with PSA <1 ng/mL (data not shown), associations were generally similar to when restricting to men with PSA <2 ng/mL, including for rs3212227.
DISCUSSION
Our prospective study supports the conclusion that variation in some genes involved in inflammation and the immune response may be associated with prostate cancer risk, and for some variants, possibly with lower- or higher-grade disease. We investigated 16 candidate SNPs in 12 genes for their association with prostate cancer, disease grade, and PSA concentration. We observed a positive association for the minor allele of rs3212227 in IL12(p40) with prostate cancer, especially lower-grade disease. We also noted a possible inverse association for the minor allele of rs486907 in RNASEL and total prostate cancer, and for rs4073 in IL8 and higher-grade disease. Two of seven tagSNPs in IL10 were associated with prostate cancer or grade of disease (rs3021094, rs1800890); the haplotype analysis did not provide additional information. None of the other candidates (IL2, IL6, IL1β, IL4, IFNG, MSR1, TLR4, TNFA) or IL10 tagSNPs was significantly associated with risk overall or by grade.
IL-12 is a TH1 cytokine, associated with anti-tumor T cell function, and IL-12 has been studied as an anti-cancer therapy both in animal models of cancer and in humans (15–17). Thus, it seems logical to hypothesize that decreased IL-12 levels would be associated with an increased risk of prostate cancer. Indeed, in this study, we found that the minor allele (C) of rs321227 in IL12(p40), which produces less of the p40 subunit of IL-12 was positively associated with risk of prostate cancer overall and lower-grade disease, even when restricting the analyses to men with low PSA. To our knowledge, these data are novel and are consistent with a previous study and a meta-analysis reporting that the C allele of rs321227 was associated with an increased risk of cancer at other sites and of cancer overall (18,19). It should be noted, however, that these findings are not consistent with other data suggesting an association of the C allele of rs321227 with a decreased cancer risk (14,20,21) or reporting a lack of association (22,23).
IL10 was a particular focus of this study because of our team’s prior findings that promoter and tagSNPs in this gene were associated with incident [12,13] and recurrent prostate cancer [12,13]. In our previous studies, the G allele of rs1800896, known to produce increased levels of this anti-inflammatory cytokine, was inversely associated with both risk of prostate cancer [12] and risk of recurrence [13]. Consistent with those data, the A allele of rs1800872, which results in decreased IL-10 levels, was positively associated the risk of prostate cancer recurrence [13] but not significantly associated with the risk of developing prostate cancer [12]. TagSNPs rs1800890 and rs3024496 were associated with both incidence (rs1800890: A inverse, rs3024496: C positive) [12] and recurrence (rs1800890: A inverse, rs3024496 C inverse) [13], and tagSNPs rs3024498 (G inverse) and rs1800894 (T positive) were associated with recurrence [13]. In our current study on incidence, we did not observe associations for the candidate SNPs and prostate cancer overall or by grade, except when restricting to men with low PSA (performed to minimize the likelihood of detection bias). In men with lower PSA levels, the candidates rs1800871 (T allele, positive), rs1800872 (A allele, positive), rs1800896 (G allele, inverse), and tagSNPs rs1800890 (A allele, inverse), rs3021094 (C allele, positive), and rs1554286 (T allele, positive, and rs3024496 (C allele, inverse) appeared to be associated with risk overall and lower-grade disease. Two other studies have reported a lack of association of rs1800871 and rs1800872, which are in strong linkage disequilibrium, with prostate cancer [14,15], yet others reported associations with overall incidence and the incidence of high-grade disease (but not low-grade) [16] or recurrence [17]. For rs1800896, a few studies reported positive associations for the G allele and prostate cancer risk [18–20], while others reported no association [21,22]. Consistent with a meta-analysis [21], in our current study, we did not find an association between tagSNPs rs3024498, rs3024496, rs1800894 and total prostate cancer or grade of disease, including when we restricted to men with low PSA. We also reported no association between tagSNP rs3024509 and overall prostate cancer and grade of disease, which is in agreement with two previous studies [23,24]. In our prior studies we reported a null association between the minor allele (C) of rs3021094 in IL10 and prostate cancer risk and recurrence (12,13), while in the current study, we found a positive association between this SNP and total prostate cancer and PSA. While the IL10 SNPs that have been observed to be associated with incidence and/or recurrence are not fully consistent across studies, nevertheless, when taken together, these studies point to an overall protective role for IL-10 in prostate cancer.
We also observed possible inverse associations for rs4073 in IL8 and rs486907 in RNASEL with high-grade disease. Some [25,26], but not all [12,27–29] studies have reported that carriers of the minor allele (A) of rs4073, which is associated with higher production of IL8, have a decreased risk of prostate cancer [25,26]. Some studies have observed associations between the minor allele (A) of rs486907, which has lower enzymatic levels of RNASEL [30], and prostate cancer risk overall or by stage or grade [31–33]; other studies reported no association[34–38].
We were concerned that the observed SNP-prostate cancer associations we observed could be explained by the influence of SNPs involved in inflammation and the immune response on serum PSA concentration. Prior studies have found that some SNPs in a variety of genes and some SNPs identified in genomewide-association studies that are not in known regulatory or coding regions of genes are associated with circulating PSA concentration [9,10,39]. In addition, in our prior study in the PCPT, we observed that men who had more inflammation in the benign prostate tissue had higher PSA concentrations, including in controls who did not have an indication for biopsy [2]. Indeed, in the present study, we noted that PSA concentration was higher in controls with the minor allele of rs2069762 in IL2, and lower in controls with the minor alleles of rs1800795 (C) and rs1800797 (A) of IL6. PSA concentration was also higher in controls with the minor allele for two of the IL10 tagSNPs (rs3021094, rs1800894), but not for the three candidate IL10 SNPs. In contrast to our finding, a study reported higher PSA concentration in men with the minor allele (C) of rs1800795 in IL6 [40].
While the above observations suggest the possibility of detection bias when studying SNPs and prostate cancer risk, when we restricted our analyses to men with PSA <2 ng/mL (i.e., those men in whom detection bias resulting from the link between SNPs and elevated PSA, an indication for biopsy, is was less likely), rs2069762 in IL2, rs1800795 and rs1800797 in IL6, and tagSNP rs1800894 in IL10 all remained unrelated to prostate cancer risk. The higher PSA concentrations for the minor allele of tagSNP rs3021094 in IL10 did not appear to explain its positive association with total, lower-, and higher-grade prostate cancer as the association was similar in men with low PSA and in all men. Also, the association between rs3212227 in IL12(p40) and total and lower-grade prostate cancer could not be explained by PSA detection bias because this SNP was not associated with PSA concentration. While not explained by an association between these SNPs and PSA, two (rs1800871, rs1800872) of the three candidate SNPs in IL10 that were not associated with total, lower-, or higher-grade disease in all men, appeared to be positively associated with total and lower-grade disease in men with low PSA. We do not have an explanation for this difference. While PSA-associated detection bias does remain a theoretical possibility [41], empirically, the influence of these immune-related SNPs on serum PSA did not result in detection bias.
It is essential to discuss the strengths and limitations of this study. First, this was a prospective study in which the controls were sampled from the same source population as the cases, helping to ensure that the genotype frequencies in the controls reflect those expected in the population that gave rise to the cases. SNPs we selected were hypothesis- driven, thus and given the prior expectation of association, we did not systematically correct for multiple testing as might be done for agnostic genetic studies. Nevertheless when correcting for multiple testing using the conservative Bonferroni correction (0.05/23 SNPs tested = 0.0022), using the main results (Table II), only the association for rs3212227 in IL12(p40) (p=0.0017) would remain statistically significant. We used a candidate gene approach, and selected a small number of known or purported functional SNPs. Another limitation in these results is that we did not attempt to cover the variation across each of the genes and their regulatory regions, with the exception of IL10. While we observed some interesting SNP-prostate cancer associations, we could have missed others or underestimated the magnitude of associations given that SNPs in genes involved in inflammation and response to infection may only be associated with prostate cancer if men have had exposures that elicit an inflammatory response. Genotyping success was lower for DNA extracted from serum than from buffy coat, but we did not observe differences in HWE for DNA extracted from these two sources. A key finding from this study was for rs3212227 in IL12(p40), a SNP for which the test for HWE was statistically significant; we confirmed that the deviation from the expected was minor. We were uniquely able to evaluate the association between SNPs and prostate cancer in a setting in which PSA-associated detection bias is less likely to be operating because of the design of the PCPT. However, because of the combination of the trial criteria and the annual screening for prostate cancer, the cases that were diagnosed were uniformly early stage and so we were not able to evaluate associations for lethal or fatal disease. However, we were able to report on higher-grade disease, which has a poorer prognosis. Another strength of these data is that Gleason grading was performed centrally for the trial, limiting error in pathologic classification. While the overall sample size was large, for some SNPs for which the minor allele was less common, estimates were less stable in the subgroups analyses. We could not study these associations by race/ethnicity because too few of the PCPT participants were non-white. In this present study our aim was to address the link between genes involved in the immune response and prostate cancer risk. Associations were modest and thus, are unlikely to influence treatment decision making or clinical risk prediction for prostate cancer
Conclusion
In conclusion, our study suggests that variation in some inflammation and immune response genes may be associated with prostate cancer risk. These associations were not fully explained by PSA-associated detection bias. Thus, our findings add support for the role of inflammation in the etiology of prostate cancer.
Supplementary Material
Acknowledgments
Funding: This work was funded by Public Health Service grants P01 CA108964 (IM Thompson, Project 4 EA Platz), U10 CA37429 (CD Blanke), UM1 CA182883 (IM Thompson/CM Tangen), P30 CA054174 (IM Thompson), P30 CA006973 (WG Nelson), and T32 CA009314 (EA Platz) from the National Cancer Institute, National Institutes of Health.
Footnotes
A SWOG-Coordinated Study S9217
Disclosure Statement: The authors declare that they have no competing financial interests related to this paper.
The content of this work is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
References
- 1.Sfanos KS, De Marzo AM. Prostate cancer and inflammation: the evidence. Histopathology. 2012;60:199–215. doi: 10.1111/j.1365-2559.2011.04033.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gurel B, Lucia MS, Thompson IM, Goodman PJ, Tangen CM, Kristal AR, Parnes HL, Hoque A, Lippman SM, Sutcliffe S, Peskoe SB, Drake CG, Nelson WG, De Marzo AM, Platz EA. Chronic inflammation in benign prostate tissue is associated with high-grade prostate cancer in the placebo arm of the prostate cancer prevention trial. Cancer Epidemiol Biomarkers Prev. 2014;23:847–56. doi: 10.1158/1055-9965.EPI-13-1126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Thompson IM, Goodman PJ, Tangen CM, Lucia MS, Miller GJ, Ford LG, Lieber MM, Cespedes RD, Atkins JN, Lippman SM, Carlin SM, Ryan A, Szczepanek CM, Crowley JJ, Coltman CA. The influence of finasteride on the development of prostate cancer. N Engl J Med. 2003;349:215–24. doi: 10.1056/NEJMoa030660. [DOI] [PubMed] [Google Scholar]
- 4.Zheng SL, Liu W, Wiklund F, Dimitrov L, Bälter K, Sun J, Adami H-O, Johansson J-E, Sun J, Chang B, Loza M, Turner AR, Bleecker ER, Meyers DA, Carpten JD, Duggan D, Isaacs WB, Xu J, Grönberg H. A comprehensive association study for genes in inflammation pathway provides support for their roles in prostate cancer risk in the CAPS study. Prostate. 2006;66:1556–64. doi: 10.1002/pros.20496. [DOI] [PubMed] [Google Scholar]
- 5.Zheng SL, Sun J, Wiklund F, Smith S, Stattin P, Li G, Adami H-O, Hsu F-C, Zhu Y, Bälter K, Kader AK, Turner AR, Liu W, Bleecker ER, Meyers DA, Duggan D, Carpten JD, Chang B-L, Isaacs WB, Xu J, Grönberg H. Cumulative association of five genetic variants with prostate cancer. N Engl J Med. 2008;358:910–9. doi: 10.1056/NEJMoa075819. [DOI] [PubMed] [Google Scholar]
- 6.Kim S-T, Cheng Y, Hsu F-C, Jin T, Kader AK, Zheng SL, Isaacs WB, Xu J, Sun J. Prostate cancer risk-associated variants reported from genome-wide association studies: meta-analysis and their contribution to genetic Variation. Prostate. 2010;70:1729–38. doi: 10.1002/pros.21208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Liu H, Wang B, Han C. Meta-analysis of genome-wide and replication association studies on prostate cancer. Prostate. 2011;71:209–24. doi: 10.1002/pros.21235. [DOI] [PubMed] [Google Scholar]
- 8.Ahn J, Berndt SI, Wacholder S, Kraft P, Kibel AS, Yeager M, Albanes D, Giovannucci E, Stampfer MJ, Virtamo J, Thun MJ, Feigelson HS, Cancel-Tassin G, Cussenot O, Thomas G, Hunter DJ, Fraumeni JF, Hoover RN, Chanock SJ, Hayes RB. Variation in KLK genes, prostate-specific antigen and risk of prostate cancer. Nat Genet. 2008;40:1032–4. doi: 10.1038/ng0908-1032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Wiklund F, Zheng SL, Sun J, Adami H-O, Lilja H, Hsu F-C, Stattin P, Adolfsson J, Cramer SD, Duggan D, Carpten JD, Chang B-L, Isaacs WB, Grönberg H, Xu J. Association of reported prostate cancer risk alleles with PSA levels among men without a diagnosis of prostate cancer. Prostate. 2009;69:419–27. doi: 10.1002/pros.20908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Gudmundsson J, Besenbacher S, Sulem P, Gudbjartsson DF, Olafsson I, Arinbjarnarson S, Agnarsson BA, Benediktsdottir KR, Isaksson HJ, Kostic JP, Gudjonsson SA, Stacey SN, Gylfason A, Sigurdsson A, Holm H, Bjornsdottir US, Eyjolfsson GI, Navarrete S, Fuertes F, Garcia-Prats MD, Polo E, Checherita IA, Jinga M, Badea P, Aben KK, Schalken JA, van Oort IM, Sweep FC, Helfand BT, Davis M, Donovan JL, Hamdy FC, Kristjansson K, Gulcher JR, Masson G, Kong A, Catalona WJ, Mayordomo JI, Geirsson G, Einarsson GV, Barkardottir RB, Jonsson E, Jinga V, Mates D, Kiemeney LA, Neal DE, Thorsteinsdottir U, Rafnar T, Stefansson K. Genetic correction of PSA values using sequence variants associated with PSA levels. Sci Transl Med. 2010;2:62–92. doi: 10.1126/scitranslmed.3001513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Goodman PJ, Tangen CM, Kristal AR, Thompson IM, Lucia MS, Platz EA, Figg WD, Hoque A, Hsing A, Neuhouser ML, Parnes HL, Reichardt JKV, Santella RM, Till C, Lippman SM. Transition of a clinical trial into translational research: the prostate cancer prevention trial experience. Cancer Prev Res (Phila) 2010;3:1523–33. doi: 10.1158/1940-6207.CAPR-09-0256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wang M-H, Helzlsouer KJ, Smith MW, Hoffman-Bolton JA, Clipp SL, Grinberg V, De Marzo AM, Isaacs WB, Drake CG, Shugart YY, Platz EA. Association of IL10 and Other immune response- and obesity-related genes with prostate cancer in CLUE II. Prostate. 2009;69:874–85. doi: 10.1002/pros.20933. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Dluzniewski PJ, Wang M-H, Zheng SL, De Marzo AM, Drake CG, Fedor HL, Partin AW, Han M, Fallin MD, Xu J, Isaacs WB, Platz EA. Variation in IL10 and other genes involved in the immune response and in oxidation and prostate cancer recurrence. Cancer Epidemiol Biomarkers Prev. 2012;21:1774–82. doi: 10.1158/1055-9965.EPI-12-0458. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Shao N, Xu B, Mi Y, Hua L. IL-10 polymorphisms and prostate cancer risk: a meta-analysis. Prostate Cancer Prostatic Dis. 2011;14:129–35. doi: 10.1038/pcan.2011.6. [DOI] [PubMed] [Google Scholar]
- 15.Yu Z, Liu Q, Huang C, Wu M, Li G. The interleukin 10 -819C/T polymorphism and cancer risk: a HuGE review and meta-analysis of 73 studies including 15,942 cases and 22,336 controls. OMICS. 2013;17:200–14. doi: 10.1089/omi.2012.0089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Faupel-Badger JM, Kidd LCR, Albanes D, Virtamo J, Woodson K, Tangrea JA. Association of IL-10 polymorphisms with prostate cancer risk and grade of disease. Cancer Causes Control. 2008;19:119–24. doi: 10.1007/s10552-007-9077-6. [DOI] [PubMed] [Google Scholar]
- 17.Lin H-C, Liu C-C, Kang W-Y, Yu C-C, Wu TT, Wang J-S, Wu W-J, Huang C-H, Wu M-T, Huang S-P. Influence of cytokine gene polymorphisms on prostate-specific antigen recurrence in prostate cancer after radical prostatectomy. Urol Int. 2009;83:463–70. doi: 10.1159/000251189. [DOI] [PubMed] [Google Scholar]
- 18.Omrani MD, Bazargani S, Bageri M. Interlukin-10, Interferon-γ and Tumor Necrosis Factor-α Genes Variation in Prostate Cancer and Benign Prostatic Hyperplasia. Curr Urol. 2008;2:175–80. [Google Scholar]
- 19.Zabaleta J, Lin H-Y, Sierra RA, Hall MC, Clark PE, Sartor OA, Hu JJ, Ochoa AC. Interactions of cytokine gene polymorphisms in prostate cancer risk. Carcinogenesis. 2008;29:573–8. doi: 10.1093/carcin/bgm277. [DOI] [PubMed] [Google Scholar]
- 20.Liu J, Song B, Bai X, Liu W, Li Z, Wang J, Zheng Y, Wang Z. Association of genetic polymorphisms in the interleukin-10 promoter with risk of prostate cancer in Chinese. BMC Cancer. 2010;10:456. doi: 10.1186/1471-2407-10-456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zou Y-F, Wang F, Feng X-L, Tian Y-H, Tao J-H, Pan F-M, Huang F. Lack of association of IL-10 gene polymorphisms with prostate cancer: evidence from 11,581 subjects. Eur J Cancer. 2011;47:1072–9. doi: 10.1016/j.ejca.2010.11.034. [DOI] [PubMed] [Google Scholar]
- 22.Wang J, Ding Q, Shi Y, Cao Q, Qin C, Zhu J, Chen J, Yin C. The interleukin-10-1082 promoter polymorphism and cancer risk: a meta-analysis. Mutagenesis. 2012;27:305–12. doi: 10.1093/mutage/ger078. [DOI] [PubMed] [Google Scholar]
- 23.Xu J, Lowey J, Wiklund F, Sun J, Lindmark F, Hsu F-C, Dimitrov L, Chang B, Turner AR, Liu W, Adami H-O, Suh E, Moore JH, Zheng SL, Isaacs WB, Trent JM, Grönberg H. The interaction of four genes in the inflammation pathway significantly predicts prostate cancer risk. Cancer Epidemiol Biomarkers Prev. 2005;14:2563–8. doi: 10.1158/1055-9965.EPI-05-0356. [DOI] [PubMed] [Google Scholar]
- 24.Stark JR, Wiklund F, Grönberg H, Schumacher F, Sinnott JA, Stampfer MJ, Mucci LA, Kraft P. Toll-like receptor signaling pathway variants and prostate cancer mortality. Cancer Epidemiol Biomarkers Prev. 2009;18:1859–63. doi: 10.1158/1055-9965.EPI-08-0981. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.McCarron SL, Edwards S, Evans PR, Gibbs R, Dearnaley DP, Dowe A, Southgate C, Easton DF, Eeles RA, Howell WM. Influence of Cytokine Gene Polymorphisms on the Development of Prostate Cancer. Cancer Res. 2002;62:3369–72. [PubMed] [Google Scholar]
- 26.Wang N, Zhou R, Wang C, Guo X, Chen Z, Yang S, Li Y. -251 T/A polymorphism of the interleukin-8 gene and cancer risk: a HuGE review and meta-analysis based on 42 case-control studies. Mol Biol Rep. 2012;39:2831–41. doi: 10.1007/s11033-011-1042-5. [DOI] [PubMed] [Google Scholar]
- 27.Yang HP, Woodson K, Taylor PR, Pietinen P, Albanes D, Virtamo J, Tangrea JA. Genetic variation in interleukin 8 and its receptor genes and its influence on the risk and prognosis of prostate cancer among Finnish men in a large cancer prevention trial. Eur J Cancer Prev. 2006;15:249–53. doi: 10.1097/01.cej.0000199504.07947.e7. [DOI] [PubMed] [Google Scholar]
- 28.Michaud DS, Daugherty SE, Berndt SI, Platz EA, Yeager M, Crawford ED, Hsing A, Huang W-Y, Hayes RB. Genetic polymorphisms of interleukin-1B (IL-1B), IL-6, IL-8, and IL-10 and risk of prostate cancer. Cancer Res. 2006;66:4525–30. doi: 10.1158/0008-5472.CAN-05-3987. [DOI] [PubMed] [Google Scholar]
- 29.Zhang J, Dhakal IB, Lang NP, Kadlubar FF. Polymorphisms in inflammatory genes, plasma antioxidants, and prostate cancer risk. Cancer Causes Control. 2010;21:1437–44. doi: 10.1007/s10552-010-9571-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Casey G, Neville PJ, Plummer SJ, Xiang Y, Krumroy LM, Klein EA, Catalona WJ, Nupponen N, Carpten JD, Trent JM, Silverman RH, Witte JS. RNASEL Arg462Gln variant is implicated in up to 13% of prostate cancer cases. Nat Genet. 2002;32:581–3. doi: 10.1038/ng1021. [DOI] [PubMed] [Google Scholar]
- 31.Rennert H, Zeigler-Johnson CM, Addya K, Finley MJ, Walker AH, Spangler E, Leonard DGB, Wein A, Malkowicz SB, Rebbeck TR. Association of susceptibility alleles in ELAC2/HPC2, RNASEL/HPC1, and MSR1 with prostate cancer severity in European American and African American men. Cancer Epidemiol Biomarkers Prev. 2005;14:949–57. doi: 10.1158/1055-9965.EPI-04-0637. [DOI] [PubMed] [Google Scholar]
- 32.Agalliu I, Leanza SM, Smith L, Trent JM, Carpten JD, Bailey-Wilson JE, Burk RD. Contribution of HPC1 (RNASEL) and HPCX variants to prostate cancer in a founder population. Prostate. 2010;70:1716–27. doi: 10.1002/pros.21207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Fesinmeyer MD, Kwon EM, Fu R, Ostrander EA, Stanford JL. Genetic variation in RNASEL and risk for prostate cancer in a population-based case-control study. Prostate. 2011;71:1538–47. doi: 10.1002/pros.21370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Breyer JP, McReynolds KM, Yaspan BL, Bradley KM, Dupont WD, Smith JR. Genetic variants and prostate cancer risk: candidate replication and exploration of viral restriction genes. Cancer Epidemiol Biomarkers Prev. 2009;18:2137–44. doi: 10.1158/1055-9965.EPI-08-1223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Meyer MS, Penney KL, Stark JR, Schumacher FR, Sesso HD, Loda M, Fiorentino M, Finn S, Flavin RJ, Kurth T, Price AL, Giovannucci EL, Fall K, Stampfer MJ, Ma J, Mucci LA. Genetic variation in RNASEL associated with prostate cancer risk and progression. Carcinogenesis. 2010;31:1597–603. doi: 10.1093/carcin/bgq132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Zhang L-F, Mi Y-Y, Qin C, Wang Y, Cao Q, Wei J-F, Zhou Y-J, Feng N-H, Zhang W. RNASEL -1385G/A polymorphism and cancer risk: a meta-analysis based on 21 case-control studies. Mol Biol Rep. 2011;38:5099–105. doi: 10.1007/s11033-010-0657-2. [DOI] [PubMed] [Google Scholar]
- 37.San Francisco IF, Rojas PA, Torres-Estay V, Smalley S, Cerda-Infante J, Montecinos VP, Hurtado C, Godoy AS. Association of RNASEL and 8q24 variants with the presence and aggressiveness of hereditary and sporadic prostate cancer in a Hispanic population. J Cell Mol Med. 2014;18:125–33. doi: 10.1111/jcmm.12171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Zhang L-L, Sun L, Zhu X-Q, Xu Y, Yang K, Yang F, Yang Y-G, Chen G-Q, Fu J-C, Zheng C-G, Li Y, Mu X-Q, Shi X-H, Zhao F, Wang F, Yang Z, Wang B-Y. rs10505474 and rs7837328 at 8q24 cumulatively confer risk of prostate cancer in Northern Han Chinese. Asian Pac J Cancer Prev. 2014;15:3129–32. doi: 10.7314/apjcp.2014.15.7.3129. [DOI] [PubMed] [Google Scholar]
- 39.Xue WM, Coetzee GA, Ross RK, Irvine R, Kolonel L, Henderson BE, Ingles SA. Genetic determinants of serum prostate-specific antigen levels in healthy men from a multiethnic cohort. Cancer Epidemiol Biomarkers Prev. 2001;10:575–9. [PubMed] [Google Scholar]
- 40.Tan D, Wu X, Hou M, Lee SO, Lou W, Wang J, Janarthan B, Nallapareddy S, Trump DL, Gao AC. Interleukin-6 polymorphism is associated with more aggressive prostate cancer. J Urol. 2005;174:753–6. doi: 10.1097/01.ju.0000168723.42824.40. [DOI] [PubMed] [Google Scholar]
- 41.Dluzniewski PJ, Xu J, Ruczinski I, Isaacs WB, Platz EA. Polymorphisms influencing prostate-specific antigen concentration may bias genome-wide association studies on prostate cancer. Cancer Epidemiol Biomarkers Prev. 2015;24:88–93. doi: 10.1158/1055-9965.EPI-14-0863. [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.