Chronic intraprostatic inflammation is suspected to play a role in the pathogenesis of prostate cancer. Polymorphisms in cytokine genes can influence inflammation and immune response and may be related to the risk of prostate cancer. Four common single nucleotide polymorphisms (SNPs) in the genes encoding interleukin-1B (IL-1B), IL-6, and IL-8 were assessed in 503 prostate cancer cases and 652 controls, and three SNPs in IL-10 were assessed in an additional 817 prostate cancer cases and 1,190 controls (for a total of 1,320 prostate cancer cases and 1,255 controls). Cases and controls were selected from the on-going Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial and were frequency matched on age, ethnicity, time-period since initial screening, and date of blood draw. Single-locus analyses were conducted using conditional logistic regression. In addition, we did a haplotype analysis for the three IL-10 SNPs tested. Overall, no associations were detected between the seven polymorphisms in the four cytokine genes examined in this study and prostate cancer risk. Further stratifying by use of nonsteroidal anti-inflammatory drugs did not modify the associations. Findings were similar for early or advanced prostate cancers. Similarly, we observed no association between the major IL-10 haplotypes and the risk of prostate cancer. At least seven common polymorphisms in genes of inflammatory cytokines IL-1B, IL-6, IL-8, and IL-10 do not seem to play a role in the risk of prostate cancer. (Cancer Res 2006; 66(8): 4525-30)

Prostate cancer is the most common cancer among men in the United States (excluding nonmelanoma skin cancer), accounting for 33% of newly diagnosed cancers (1). Twin studies suggest that up to 50% of the risk of prostate cancer may be explained by genetic factors (2). Although high penetrance genes may explain only a small percentage of all prostate cancer cases, common genetic polymorphisms are likely to play an important role among sporadic cases (3).

Chronic inflammation may lead to tumorigenesis by damaging DNA through radical oxygen and nitrogen species, enhancing cell proliferation, and stimulating angiogenesis (4). In epidemiologic studies, sexually transmitted infections and prostatitis have been associated with prostate cancer risk in some studies (5, 6). Furthermore, nonsteroidal anti-inflammatory drugs (NSAID) seem to decrease risk of prostate cancer. In a meta-analysis of five prospective studies, aspirin use was associated with a 15% reduction in prostate cancer [relative risk (RR), 0.85; 95% confidence interval (95% CI), 0.77-0.94], and the pooled RR of four studies with information on any NSAID use was 0.67 (95% CI, 0.37-1.22; ref. 7).

Common polymorphisms in proinflammatory and anti-inflammatory cytokine genes can influence cytokine production and may play a role in prostate cancer. To date, five case-control studies have examined genetic polymorphisms in different inflammatory cytokines [interleukin-1B (IL-1B), IL-6, IL-8, IL-10, tumor necrosis factor-α (TNF-α), and macrophage-inhibitory cytokine-1 (MIC-1)] and prostate cancer risk (812). In the largest case-control study to date (1,383 prostate cancer cases), no association was observed for variants of the IL-6 gene (10), but in the same population, one of the tagging single nucleotide polymorphisms (SNPs) in the MIC-1 gene was associated with risk of prostate cancer (8). In another study (247 cases), two SNP polymorphisms in the promoter region of the genes that encode IL-8 (−251) and IL-10 (−1082) were associated with the risk of prostate cancer (12). Findings for polymorphisms in the TNF-α gene have been inconsistent (9, 11, 12).

To further investigate the possible role of genetic susceptibility within inflammatory pathways, we examined seven common genetic polymorphisms of cytokines IL-1B, IL-6, IL-8, and IL-10 in relation to prostate cancer, using a prospective study design.

Study Population

The Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial is an ongoing randomized community-based study evaluating the effectiveness of cancer screening tests on site-specific cancer mortality. Over 154,000 men and women ages 55 to 74 years were enrolled at 10 screening centers nationwide between September 1993 and July 2001. Participants in the intervention arm undergo periodic screening tests, including prostate-specific antigen (PSA, annually, years 0-5 among men) and digital rectal exam (DRE, annually, years 0-3 among men), chest X-ray, 60-cm flexible sigmoidoscopy, cancer antigen 125 (among women) blood test, and vaginal ultrasound; more details on screening are described elsewhere (13). Participants randomized to the control arm follow their usual medical care practices. (Control arm participants are not considered in the present investigation.) Follow-up in both groups will take place for at least 13 years from randomization to assess health status and cause of death. The 10 primary screening centers are located at Washington, DC; Detroit, MI; Salt Lake City, UT; Denver, CO; Honolulu, HI; Minneapolis, MN; Marshfield, WI; Pittsburgh, PA; St. Louis, MO; and Birmingham, AL (for details on collaborating institutions, see ref. 14).

The eligibility criteria for the PLCO trial (pertinent to this analysis) were (a) age (55-75 years); (b) no current treatment for cancer except basal cell or squamous cell skin cancer; (c) no known prior cancer of the colon, rectum, prostate, lung, or ovaries; (d) no surgical removal of the colon, lung, ovary, or prostate; and (e) no more than one PSA test in past 3 years (for men randomized after April 1995); other criteria not relevant to this analysis are cited elsewhere (13).

At baseline, data on demographic characteristics, medical history, family history of cancer, screening history, and known risk factors for the cancers under study were collected from all participants. Blood samples were obtained at baseline and in subsequent screening exams from participants in the screening arm (15). All samples were shipped overnight to a central biorepository and were stored at −70°C. In the screening arm, ∼82% of the participants provided at least one blood sample.

Institutional review boards at each of the participating institutions approved the PLCO protocol, and participants provided written informed consent.

Case Ascertainment

Prostate cancer cases in the screening arm are identified through several activities. A positive screening test for prostate cancer (PSA > 4 ng/mL or a DRE suspicious for cancer) results in a referral for clinical work-up (which generally includes a biopsy). Participants also report the diagnosis of cancer on annual mailed questionnaires asking about type and date of cancer diagnosed in the previous year. Participants who do not respond to the annual questionnaire are contacted by repeat mailings or telephone. In addition, the National Death Index is periodically searched to enhance completeness of end point ascertainment. For all subjects with a positive screen for prostate cancer, a self-report of prostate cancer (on annual questionnaire), or other reports that suggest prostate cancer, trained medical record abstractors recorded medical and pathologic data pertinent to the diagnosis of prostate cancer, including date of diagnosis, stage, and grade (Gleason sum). We considered “advanced” cases to be those with aggressive potential (i.e., those that would be most likely to progress). A case was considered “advanced” if it was diagnosed with extraprostatic extension or metastasis (stages III or IV) or with Gleason sum of ≥7.

Selection of Cases and Matching Controls

White non-Hispanic and Black non-Hispanic men were eligible for study if they were randomized to the screening arm, had a valid PLCO screening for prostate cancer (PSA or DRE) between October 1993 and September 2001, completed the baseline risk factor questionnaire, provided a blood sample, and signed the informed consent for studies of cancer (n = 28,243). The eligible group included 1,320 prostate cancer cases (1,213 non-Hispanic Whites and 107 non-Hispanic Blacks). For comparison, 1,842 men (Whites = 1,433, Blacks = 409) without a diagnosis of prostate cancer at the time the cases were diagnosed, were selected as controls using incidence density sampling, and frequency matched by age (55-59, 60-64, 65-69, 70-74 years), ethnicity (1.2:1 for Whites, 4:1 for Blacks), time since initial screening (1-year periods), and date of blood draw (1-year calendar periods). All subjects were followed from their first valid prostate cancer screen to first occurrence of prostate cancer, loss-to-follow-up, death, or September 30, 2001, whichever came first.

To efficiently use study resources, we measured all the polymorphisms in roughly half of the total sample size (phase I) and only measured SNPs in the other half of the sample (phase II) if the results from phase I suggested that there may be an association. In phase I, we oversampled advanced (high stage or high grade) prostate cancer cases and randomly selected controls frequency matched by the same four factors as the full study. All SNPs were measured in 503 prostate cancer cases and 652 matched controls (phase I). The IL-10 SNPs were measured in both phase I and II (817 cases and 1,190 controls). Because of the incidence density sampling, 105 controls were selected to serve more than once as a control, and 72 controls subsequently became cases in the total sample. Cases are considered prevalent if they were diagnosed with prostate cancer within 1 year of the first screening (n = 212, phase I; n = 241, phase II) and incident if the diagnosis occurred anytime thereafter (n = 291, phase I; n = 576, phase II).

Cytokine Gene Polymorphisms

All DNA extractions and genotyping for this study were done at the Core Genotyping Facility at the National Cancer Institute, Frederick, MD. A Taqman platform was used for genotyping (Applied Biosystems, Foster City, CA). Detailed information of primers, probes, and conditions used in the genotyping assays is available in the National Cancer Institute's SNP500Cancer Database (16).8

Seven SNPs were studied in four genes: IL-1B, IL-6, IL-8, and IL-10. The SNPs in this study were selected from four genes of interest based on allele frequency and functional implications from previous publications and data on tagging SNPs for haplotype analysis (using available resources at the time of SNP selection). We labeled these polymorphisms using their dbSNP ID. The following SNPs were assessed: rs16944, rs1143634, rs1800795, rs4073, rs1800896, rs1800871, rs3024496. The exon SNP (rs1143634) for IL-1B is a nonsynonymous SNP (i.e., results in an amino acid change), the exon SNP (rs3024496) for IL-10 is synonymous, and the remaining SNPs are located in the promoter region of the genes.

Blinded DNA quality control samples from 48 different individuals were repeatedly placed (three to seven repeats) among the study samples with blinding of laboratory personnel to the identity of the quality samples and case-control status. Interassay concordance was 96.5% for IL-6 SNP but was >99% for all other SNPs. Genotype data for each SNP were successfully obtained on >93% of study subjects. Lack of genotype data was due to insufficient DNA (4.6% for phase I samples; 3.9% for phase II samples) and genotyping failure (range, 0.4-2.2%).

Exposure Information

At study entry, all participants filled out a risk factor questionnaire, including age; ethnicity; education; occupation; current and past smoking behavior; history of cancer and other diseases; use of selected drugs, including those containing aspirin and ibuprofen; recent history of cancer screening exams; and prostate-related health factors. Current NSAID use was defined as regular use of aspirin or ibuprofen in the last 12 months (use ≥4/mo to ≥60/mo), and irregular NSAID users were those individuals who did not report regular use (<4 times per month) of aspirin or ibuprofen.

Statistical Analysis

Single-locus analysis. Genotype frequency differences between cases and controls were evaluated using standard contingency X2 tests and Ps. Hardy-Weinberg equilibrium was tested among controls for each IL sequence variant using Pearson's χ2. We used conditional logistic regression to assess the association between IL genotypes and prostate cancer risk among all races and Whites only to maximize power. Odds ratios (OR) were estimated under a codominant model (homozygote and heterozygote genotypes that carry the variant allele versus homozygote wild type, the designated reference group). Tests for trend were assessed using a continuous variable for genotype. We did subgroup analyses of advanced prostate cancer. Stratification was used to evaluate whether the association between the cytokine polymorphisms and prostate cancer risk differed by NSAID use. The test for interaction was evaluated by the Wald test for the β coefficient of the cross-product term in a model that contained the main effects of the genotype and NSAIDs in addition to the interaction term.

Haplotype analysis. The haplotype analysis was conducted among Whites only as haplotype frequencies may vary substantially by race. Pairwise linkage disequilibrium tests were estimated in Haploview v2.05 for the IL-10 variants among controls. D′ (17) and r2 (18) were used to estimate the magnitude of linkage disequilibrium between the sequence variants. We used a global score test developed by Schaid et al. (19) to assess differences in overall haplotype frequency profiles between cases and controls adjusting for age, study year, and time of follow-up. Ps for the global tests were estimated from empirical distributions created from a minimum of 1,000 permuted data sets. The association between haplotypes and prostate cancer risk was estimated by regression substitution, which is an iterative process that uses posterior probabilities of haplotypes as weights to update the regression coefficients and regression coefficients to update the posterior probabilities (20), as implemented in the module haplo.glm in the computer program Haplo Stats for the R programming language.9

ORs were estimated for the association between haplotypes and prostate cancer risk under the additive model using generalized linear models adjusting for age, study year, and time of follow-up. The test for interaction was evaluated using the likelihood ratio test comparing the model with interaction terms for NSAID use and IL-10 haplotypes to the model without the interaction terms.

Genetic polymorphisms in the IL-10 gene were measured in two separate batches (phase I and II: 1,245 cases and 1,743 controls); SNPs in the remaining three cytokine genes (IL-1B, IL-6, and IL-8) were only measured in phase I subjects (503 cases and 652 controls). Table 1 provides the baseline characteristics for cases and controls in the two batches. The median PSA level (ng/mL) at diagnosis was 5.8 and 5.4 for phase I and II, respectively. Percentage of prostate cancer cases with advanced disease (stages III-IV or Gleason sum ≥ 7) was higher in phase I (59.3%) than phase II (31.7%; Table 1) because of oversampling of advanced cases in phase I. Baseline characteristics, including attained education, body mass index, vasectomy, and current use of NSAID, were similar in cases and controls and in the two batches (Table 1). In both batches, cases had a higher prevalence of a family history of prostate cancer and were slightly less likely to be current smokers than controls.

Table 1.

Baseline and clinical characteristics of prostate cancer cases and controls in the PLCO study

Phase I
Phase II
Case (n = 503)Control (n = 652)Case (n = 817)Control (n = 1,190)
Mean (SD) age, y 67.1 (5.2) 66.6 (5.3) 66.9 (5.4) 66.7 (5.4) 
Race (%)     
    White 479 (95.2) 567 (87.0) 734 (89.8) 866 (72.8) 
    Black 24 (4.8) 85 (13.0) 83 (10.2) 324 (27.2) 
Education (%)     
    Less than high school 54 (10.8) 58 (8.9) 84 (10.3) 108 (9.1) 
    High school 88 (17.5) 105 (16.1) 141 (17.3) 245 (20.6) 
    Vocational 161 (32.1) 235 (36.0) 250 (30.6) 378 (31.8) 
    College 199 (39.6) 254 (39.0) 341 (41.8) 457 (38.5) 
Family history of prostate cancer (%)     
    Yes 53 (10.9) 45 (7.2) 90 (11.4) 83 (7.1) 
    No 433 (89.1) 580 (92.8) 702 (88.6) 1,083 (92.9) 
Gonorrhea (%)     
    Yes 29 (5.8) 41 (6.3) 48 (5.9) 132 (11.4) 
    No 468 (94.2) 606 (93.7) 760 (94.1) 1,030 (88.6) 
Syphilis (%)     
    Yes 2 (0.4) 10 (1.6) 6 (0.7) 21 (1.8) 
    No 494 (99.6) 627 (98.4) 801 (99.3) 1,133 (98.2) 
NSAID use (%)     
    Yes 235 (46.7) 334 (51.2) 408 (49.9) 595 (50.0) 
    No 268 (53.3) 318 (48.8) 409 (50.1) 595 (50.0) 
Smoking (%)     
    Never 178 (35.5) 171 (26.2) 248 (30.4) 377 (31.7) 
    Current 42 (8.4) 72 (11.0) 75 (9.2) 140 (11.8) 
    Ex smoker 245 (48.8) 351 (53.8) 421 (51.6) 580 (48.7) 
    Pipe/Cigar 37 (7.4) 58 (8.9) 72 (8.8) 93 (7.8) 
BMI in kg/m2 (%)     
    0 ≤ BMI < 18.5  3 (0.5) 1 (0.1) 9 (0.8) 
    18.5 ≤ BMI < 25 136 (27.4) 166 (25.7) 222 (27.6) 293 (25.0) 
    25 ≤ BMI<30 261 (52.5) 328 (50.9) 419 (52.1) 574 (48.9) 
    BMI ≥ 30 100 (20.1) 148 (23.0) 163 (20.3) 296 (25.3) 
Vasectomy (%)     
    Yes 109 (21.8) 154 (23.6) 212 (26.0) 257 (21.7) 
    No 391 (78.2) 498 (76.4) 604 (74.0) 929 (78.3) 
Advanced cancer* (%)     
    Yes 293 (59.3)  253 (31.7)  
    No 201 (40.7)  545 (68.3)  
Metastatic cancer (%)     
    MX 66 (13.3)  125 (15.4)  
    M0 425 (85.3)  672 (82.9)  
    M1A  2 (0.3)  
    M1B 7 (1.4)  12 (1.5)  
PSA (ng/mL)     
    Median (interquartile range) 5.8 (4.1-8.7) 1.2 (0.7-2.1) 5.4 (3.9-7.7) 1.2 (0.7-2.3) 
Phase I
Phase II
Case (n = 503)Control (n = 652)Case (n = 817)Control (n = 1,190)
Mean (SD) age, y 67.1 (5.2) 66.6 (5.3) 66.9 (5.4) 66.7 (5.4) 
Race (%)     
    White 479 (95.2) 567 (87.0) 734 (89.8) 866 (72.8) 
    Black 24 (4.8) 85 (13.0) 83 (10.2) 324 (27.2) 
Education (%)     
    Less than high school 54 (10.8) 58 (8.9) 84 (10.3) 108 (9.1) 
    High school 88 (17.5) 105 (16.1) 141 (17.3) 245 (20.6) 
    Vocational 161 (32.1) 235 (36.0) 250 (30.6) 378 (31.8) 
    College 199 (39.6) 254 (39.0) 341 (41.8) 457 (38.5) 
Family history of prostate cancer (%)     
    Yes 53 (10.9) 45 (7.2) 90 (11.4) 83 (7.1) 
    No 433 (89.1) 580 (92.8) 702 (88.6) 1,083 (92.9) 
Gonorrhea (%)     
    Yes 29 (5.8) 41 (6.3) 48 (5.9) 132 (11.4) 
    No 468 (94.2) 606 (93.7) 760 (94.1) 1,030 (88.6) 
Syphilis (%)     
    Yes 2 (0.4) 10 (1.6) 6 (0.7) 21 (1.8) 
    No 494 (99.6) 627 (98.4) 801 (99.3) 1,133 (98.2) 
NSAID use (%)     
    Yes 235 (46.7) 334 (51.2) 408 (49.9) 595 (50.0) 
    No 268 (53.3) 318 (48.8) 409 (50.1) 595 (50.0) 
Smoking (%)     
    Never 178 (35.5) 171 (26.2) 248 (30.4) 377 (31.7) 
    Current 42 (8.4) 72 (11.0) 75 (9.2) 140 (11.8) 
    Ex smoker 245 (48.8) 351 (53.8) 421 (51.6) 580 (48.7) 
    Pipe/Cigar 37 (7.4) 58 (8.9) 72 (8.8) 93 (7.8) 
BMI in kg/m2 (%)     
    0 ≤ BMI < 18.5  3 (0.5) 1 (0.1) 9 (0.8) 
    18.5 ≤ BMI < 25 136 (27.4) 166 (25.7) 222 (27.6) 293 (25.0) 
    25 ≤ BMI<30 261 (52.5) 328 (50.9) 419 (52.1) 574 (48.9) 
    BMI ≥ 30 100 (20.1) 148 (23.0) 163 (20.3) 296 (25.3) 
Vasectomy (%)     
    Yes 109 (21.8) 154 (23.6) 212 (26.0) 257 (21.7) 
    No 391 (78.2) 498 (76.4) 604 (74.0) 929 (78.3) 
Advanced cancer* (%)     
    Yes 293 (59.3)  253 (31.7)  
    No 201 (40.7)  545 (68.3)  
Metastatic cancer (%)     
    MX 66 (13.3)  125 (15.4)  
    M0 425 (85.3)  672 (82.9)  
    M1A  2 (0.3)  
    M1B 7 (1.4)  12 (1.5)  
PSA (ng/mL)     
    Median (interquartile range) 5.8 (4.1-8.7) 1.2 (0.7-2.1) 5.4 (3.9-7.7) 1.2 (0.7-2.3) 
*

Stage III-IV or Gleason sum ≥ 7.

All genotype distributions were consistent with Hardy-Weinberg proportions in the controls. In the single-locus analysis, IL-1B, IL-6, and IL-8 polymorphisms were not associated with the risk of prostate cancer (Table 2). Findings were similar for early and advanced stages of cancer for IL-1B and IL-6 (Table 2); among advanced cases, a slight decrease in risk was observed with the IL-8 (rs4073) variant (A) allele (Table 2), but the association was not statistically significant (trend test, P = 0.08). Associations were similar by race and when cases were stratified by “prevalent” (cancer detected in first year of follow-up) and “incident” prostate cancer (data not shown). No associations were observed for the polymorphisms in IL-1B, IL-6, and IL-8 and prostate cancer risk when stratified by NSAID use (data not shown).

Table 2.

IL-1B, IL-6, and IL-8 polymorphisms and risk of prostate cancer (overall and by stage) in the PLCO study, 1993-2001 (phase I)

Gene and dbSNP IDCasesControlsOverall, OR* (95% CI)Advanced stage (n = 124), OR* (95% CI)Early stage (n = 359), OR* (95% CI)
IL-1B      
rs16944      
    CC 218 248 1.0 1.0 1.0 
    CT 204 274 0.86 (0.66-1.12) 0.85 (0.54-1.34) 0.85 (0.64-1.13) 
    TT 51 85 0.80 (0.54-1.20) 1.03 (0.53-2.01) 0.74 (0.48-1.16) 
rs1143634      
    CC 296 375 1.0 1.0 1.0 
    CT 162 212 0.92 (0.71-1.19) 0.90 (0.58-1.39) 1.15 (0.62-2.12) 
    TT 28 27 1.22 (0.70-2.12) 1.54 (0.64-3.68) 0.91 (0.69-1.21) 
IL-6      
rs1800795      
    GG 170 230 1.0 1.0 1.0 
    CG 223 293 0.89 (0.67-1.17) 0.77 (0.48-1.21) 0.93 (0.68-1.26) 
    CC 91 90 1.18 (0.82-1.69) 1.00 (0.55-1.83) 1.21 (0.82-1.80) 
IL-8      
rs4073      
    TT 147 152 1.0 1.0 1.0 
    AT 225 310 0.78 (0.58-1.04) 0.54 (0.34-0.88) 0.89 (0.64-1.23) 
    AA 112 151 0.91 (0.65-1.29) 0.64 (0.36-1.13) 1.02 (0.70-1.48) 
Gene and dbSNP IDCasesControlsOverall, OR* (95% CI)Advanced stage (n = 124), OR* (95% CI)Early stage (n = 359), OR* (95% CI)
IL-1B      
rs16944      
    CC 218 248 1.0 1.0 1.0 
    CT 204 274 0.86 (0.66-1.12) 0.85 (0.54-1.34) 0.85 (0.64-1.13) 
    TT 51 85 0.80 (0.54-1.20) 1.03 (0.53-2.01) 0.74 (0.48-1.16) 
rs1143634      
    CC 296 375 1.0 1.0 1.0 
    CT 162 212 0.92 (0.71-1.19) 0.90 (0.58-1.39) 1.15 (0.62-2.12) 
    TT 28 27 1.22 (0.70-2.12) 1.54 (0.64-3.68) 0.91 (0.69-1.21) 
IL-6      
rs1800795      
    GG 170 230 1.0 1.0 1.0 
    CG 223 293 0.89 (0.67-1.17) 0.77 (0.48-1.21) 0.93 (0.68-1.26) 
    CC 91 90 1.18 (0.82-1.69) 1.00 (0.55-1.83) 1.21 (0.82-1.80) 
IL-8      
rs4073      
    TT 147 152 1.0 1.0 1.0 
    AT 225 310 0.78 (0.58-1.04) 0.54 (0.34-0.88) 0.89 (0.64-1.23) 
    AA 112 151 0.91 (0.65-1.29) 0.64 (0.36-1.13) 1.02 (0.70-1.48) 

NOTE: The genotype distributions were consistent with the Hardy-Weinberg equilibrium.

*

Conditional logistic regression: cases and controls were frequency matched on race, year of entry into study, time since study entry, and age.

In the two batches combined, the three IL-10 SNPs were not associated with prostate cancer risk overall, when stratified by stage, or by NSAID use (Table 3). In phase I, we originally observed statistically significant associations between each of the three IL-10 polymorphisms and prostate cancer risk among non-NSAID users (e.g., Ptrend ≤ 0.001 for SNP rs1800896), whereas no association was observed among NSAID users (all Pinteraction < 0.04; data not shown). However, the associations observed in phase I did not persist after combining data from phase II.

Table 3.

IL-10 polymorphisms and risk of prostate cancer in the PLCO study (1993-2001) overall, by stage and by NSAID use (phase I and II)

IL-10 SNPCases*Controls*Overall, OR (95% CI)Advanced stage, OR (95% CI)Early stage, OR (95% CI)NSAID use, OR (95% CI)Non-NSAID use, OR (95% CI)
rs1800896        
    AA 356 523 1.0 1.0 1.0 1.0 1.0 
    AG 599 857 0.97 (0.81-1.15) 0.75 (0.49-1.14) 1.01 (0.84-1.21) 0.77 (0.60-1.00) 1.22 (0.94-1.57) 
    GG 290 383 0.98 (0.80-1.21) 0.81 (0.57-1.15) 1.05 (0.84-1.31) 0.97 (0.72-1.31) 0.96 (0.71-1.31) 
rs1800871        
    CC 716 964 1.0 1.0 1.0 1.0 1.0 
    CT 447 659 1.01 (0.86-1.19) 1.28 (0.94-1.76) 0.96 (0.81-1.14) 1.03 (0.82-1.29) 1.02 (0.81-1.28) 
    TT 83 139 0.99 (0.74-1.34) 1.58 (0.87-2.87) 0.91 (0.66-1.25) 0.99 (0.63-1.57) 0.86 (0.56-1.32) 
rs3024496        
    TT 358 511 1.0 1.0 1.0 1.0 1.0 
    CT 599 858 0.95 (0.80-1.13) 0.76 (0.54-1.07) 1.00 (0.83-1.21) 0.78 (0.60-1.01) 1.18 (0.91-1.52) 
    CC 291 407 0.95 (0.77-1.17) 0.71 (0.47-1.08) 1.02 (0.82-1.27) 0.94 (0.70-1.27) 0.96 (0.71-1.31) 
IL-10 SNPCases*Controls*Overall, OR (95% CI)Advanced stage, OR (95% CI)Early stage, OR (95% CI)NSAID use, OR (95% CI)Non-NSAID use, OR (95% CI)
rs1800896        
    AA 356 523 1.0 1.0 1.0 1.0 1.0 
    AG 599 857 0.97 (0.81-1.15) 0.75 (0.49-1.14) 1.01 (0.84-1.21) 0.77 (0.60-1.00) 1.22 (0.94-1.57) 
    GG 290 383 0.98 (0.80-1.21) 0.81 (0.57-1.15) 1.05 (0.84-1.31) 0.97 (0.72-1.31) 0.96 (0.71-1.31) 
rs1800871        
    CC 716 964 1.0 1.0 1.0 1.0 1.0 
    CT 447 659 1.01 (0.86-1.19) 1.28 (0.94-1.76) 0.96 (0.81-1.14) 1.03 (0.82-1.29) 1.02 (0.81-1.28) 
    TT 83 139 0.99 (0.74-1.34) 1.58 (0.87-2.87) 0.91 (0.66-1.25) 0.99 (0.63-1.57) 0.86 (0.56-1.32) 
rs3024496        
    TT 358 511 1.0 1.0 1.0 1.0 1.0 
    CT 599 858 0.95 (0.80-1.13) 0.76 (0.54-1.07) 1.00 (0.83-1.21) 0.78 (0.60-1.01) 1.18 (0.91-1.52) 
    CC 291 407 0.95 (0.77-1.17) 0.71 (0.47-1.08) 1.02 (0.82-1.27) 0.94 (0.70-1.27) 0.96 (0.71-1.31) 
*

Number of cases and controls are for overall analysis. Case numbers (for SNP rs1800896, rs1800871, and rs3024496, respectively) in advanced-stage analysis: 205, 205, 206; early-stage analysis: 1034, 1035, 1036; NSAID use analysis: 605, 608, 604; non-NSAID use analysis: 640, 638, 644.

Conditional logistic regression: cases and controls were frequency matched on race, year of entry into study, time since study entry, and age.

We undertook haplotype analyses for the three IL-10 SNPs among Whites (Table 4). Linkage disequilibrium between the allelic variants rs1800896 and rs3024496 was complete (r2 = 0.99, D′ = 1). Linkage disequilibrium was not as high for the other two IL-10 SNPs pairs (r2 = 0.28 between rs1800871 and rs3024496; r2 = 0.29 between rs1800871 and rs1800896; D′ = 1 for both pairs). In our haplotype analysis, we observed three common haplotypes (>5% among controls: ATT, ACT, and GCC). The empirical P for the global score test did not indicate a statistically significant difference in haplotype frequency profiles between cases and controls overall or by NSAID use (Table 4). The A-T-T and A-C-T haplotypes were not associated with prostate cancer, overall or when stratified by NSAID use, when compared with the G-C-C (rs1800896, rs1800871, and rs3024496) haplotype (Table 4).

Table 4.

OR (95% CI) of prostate cancer for IL-10 haplotypes (rs1800896, rs1800871, and rs3024496) among Caucasians in the PLCO study (1993-2001; phase I and II)

OverallNSAID usersNon-NSAID users
Global score statistic (P  0.81 0.83 0.77 
      
Haplotype Frequencies
 
 OR (95% CI) OR (95% CI) OR (95% CI) 
 Cases (n = 1,153) Controls (n = 1,388)    
G-C-C 0.477 0.488 Reference Reference Reference 
A-T-T 0.234 0.233 1.03 (0.89-1.18) 1.04 (0.85-1.26) 1.01 (0.83-1.23) 
A-C-T 0.287 0.277 1.06 (0.93-1.21) 1.05 (0.88-1.26) 1.09 (0.90-1.32) 
Pinteraction = 0.83      
OverallNSAID usersNon-NSAID users
Global score statistic (P  0.81 0.83 0.77 
      
Haplotype Frequencies
 
 OR (95% CI) OR (95% CI) OR (95% CI) 
 Cases (n = 1,153) Controls (n = 1,388)    
G-C-C 0.477 0.488 Reference Reference Reference 
A-T-T 0.234 0.233 1.03 (0.89-1.18) 1.04 (0.85-1.26) 1.01 (0.83-1.23) 
A-C-T 0.287 0.277 1.06 (0.93-1.21) 1.05 (0.88-1.26) 1.09 (0.90-1.32) 
Pinteraction = 0.83      

NOTE: Haplotypes not shown were present in <1% of individuals. All risk estimates and score statistics are adjusted year of entry into study, time since study entry, and age.

We examined seven SNPs in four cytokine genes (IL-1B, IL-6, IL-8, and IL-10) known to be involved in the inflammatory response and previously associated with prostate cancer in experimental data. In this prospective study, we observed no association between these seven common SNPs in four cytokines of interest and prostate cancer risk. Findings were similar when prostate cancer cases were stratified by stage and grade. No interaction by NSAID use was observed for any of the SNPs examined in this study.

We selected SNPs in IL-1B, IL-6, and IL-8 as experimental data suggest that these cytokines play a role in prostate tumor progression. The IL-6 receptor correlates with proliferation of prostate cancer cells in vivo (21), and serum IL-6 is elevated in patients with metastatic prostate cancer and seems to mediate survival (22). The proinflammatory cytokines IL-1 and TNF produce powerful induction of IL-8 in a wide variety of cells. IL-8 promotes endothelial cell proliferation and chemotaxis in vitro (23) and causes neovascularization of rat and rabbit cornea in vivo (24). Two recent studies provided direct evidence that IL-8 regulates the growth and metastasis of human prostate cancer (25, 26). Elevated serum levels of IL-8 have been observed in prostate cancer patients and were shown to predict the severity of disease in one study (27).

Unlike IL-1B, IL-6, and IL-8, which are proinflammatory cytokines and part of the innate immune system, IL-10 regulates both cellular and humoral immunity. IL-10 is an anti-inflammatory cytokine that suppresses the Th-1 response but also regulates growth and/or differentiation of B cells, natural killer cells, cytotoxic, and helper T cells (28). IL-10 may play a role in prostate cancer through its ability to inhibit the production of proinflammatory cytokines, including IL-1, IL-6, and IL-8 (29). IL-10 has also been shown inhibit angiogenesis of immortalized human prostate cancer cell lines in vitro (30).

In a case-control study of 247 prostate cancer cases, a statistically significant association was reported between a polymorphism in the IL-8 gene (251A/T; same as rs4073A/T in our article) and prostate cancer risk (12). Individuals with the TT variant had a RR of 0.66 (95% CI, 0.44-0.99) compared with the AA variant. An association with risk for prostate cancer was also observed with the IL-10 polymorphism −1082 AA (same as rs1800896 in our article; OR, 1.78; 95% CI, 1.14-2.77, compared with the GG genotype; ref. 12). Both the IL-8 and the IL-10 findings in that study were inconsistent with our results. Consistent with our study, no associations were detected for the IL-1B polymorphism −511C/T (same as rs16944 in our article) and prostate cancer risk (12). The McCarron et al. study had a small sample size, and controls were unrepresentative of the cases (controls, composed of men and women, were drawn from an organ donor bank). Therefore, their results may have been a consequence of chance or bias due to control selection (organ donors tend to be healthier as they are screened for infectious diseases and consequently may have had different variant frequencies than the cases).

In a large case-control study (n = 1,383 prostate cancer cases), no associations were observed between six SNPs in the IL-6 gene and prostate cancer risk (10). Furthermore, in that study, the RR for CC versus GG of SNP rs1800795 was identical to ours (OR, 1.18; 95% CI, 0.91-1.52; ref. 10). Other studies to date on cytokine polymorphisms and prostate cancer did not include the cytokine genes presented in this article.

Although experimental data suggest that cytokines IL-1B, IL-6, IL-8, and IL-10 play a role in prostate cancer, our data, along with results from a large case-control study (10), do not suggest that several common polymorphisms in these genes are associated with prostate cancer risk. A number of possibilities exist for these inconsistencies, including the role of these genes in progression or mortality of prostate cancer versus incidence; we had limited power to examine this in our study. Alternatively, associations may be missed in epidemiologic studies because of complex gene-environment or gene-gene interactions.

The strengths of our study include a prospective design, a nested age-matched and race-matched case-control set, data on potential effect modifiers, and a large number of cases. A nested case-control design, where controls are selected from the same source population as the cases, prevents potential bias that may occur in standard case-control studies if allele frequencies in the control group are not representative of those in the population that gives rise to the cases. An additional strength of this study is the high-quality of genotyping (i.e., low genotyping failure rate and excellent reproducibility).

A potential limitation of our study is that it is a screening trial, and prostate cancers are often detected early with PSA screening. Our results may not be generalizable to men who are less likely to be screened and are diagnosed at later stages of the disease. Overall, 42% of the prostate cancer cases our study had advanced disease (stage III/IV or Gleason sum ≥ 7), of those, <2% had metastatic disease. Finally, although we did not observe associations with the selected SNPs in this study, we cannot exclude the possibility that other, unmeasured, SNPs in these genes may be associated with prostate cancer risk.

We did not observe associations between selected SNPs in four inflammatory cytokine (IL-1B, IL-6, IL-8, and IL-10) gene polymorphisms and the risk of prostate cancer. Furthermore, use of NSAIDs did not modify the observed associations, and our haplotype analysis of IL-10 (based on three SNPs) did not reveal any associations with prostate cancer risk. Overall, our findings do not suggest that the seven common polymorphisms examined are associated with the risk of prostate cancer. However, there are other SNPs in the genes investigated in this study and many genes involved in inflammation that may be related to prostate cancer risk; future studies of prostate cancer will need to thoroughly investigate promising genes in the inflammatory pathways.

Grant support: Intramural Research Program of the National Cancer Institute, NIH.

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.

We thank Drs. Christine Berg and Philip Prorok (Division of Cancer Prevention, National Cancer Institute); the Screening Center investigators and staff of the PLCO Cancer Screening Trial; Tom Riley and staff (Information Management Services, Inc.); Barbara O'Brien and staff (Westat, Inc.); and Drs. Bill Kopp, Wen Shao, and staff (Science Applications International Corporation-Frederick) for their contributions.

1
Cancer facts and figures 2005. Atlanta: American Cancer Society, Inc.; 2005.
2
Lichtenstein P, Holm NV, Verkasalo PK, et al. Environmental and heritable factors in the causation of cancer-analyses of cohorts of twins from Sweden, Denmark, and Finland.
N Engl J Med
2000
;
343
:
78
–85.
3
Schaid DJ. The complex genetic epidemiology of prostate cancer.
Hum Mol Genet
2004
;
13
Spec No 1:
R103
–21.
4
Coussens LM, Werb Z. Inflammation and cancer.
Nature
2002
;
420
:
860
–7.
5
Dennis LK, Lynch CF, Torner JC. Epidemiologic association between prostatitis and prostate cancer.
Urology
2002
;
60
:
78
–83.
6
Dennis LK, Dawson DV. Meta-analysis of measures of sexual activity and prostate cancer.
Epidemiology
2002
;
13
:
72
–9.
7
Mahmud S, Franco E, Aprikian A. Prostate cancer and use of nonsteroidal anti-inflammatory drugs: systematic review and meta-analysis.
Br J Cancer
2004
;
90
:
93
–9.
8
Lindmark F, Zheng SL, Wiklund F, et al. H6D polymorphism in macrophage-inhibitory cytokine-1 gene associated with prostate cancer.
J Natl Cancer Inst
2004
;
96
:
1248
–54.
9
Oh BR, Sasaki M, Perinchery G, et al. Frequent genotype changes at -308, and 488 regions of the tumor necrosis factor-alpha (TNF-alpha) gene in patients with prostate cancer.
J Urol
2000
;
163
:
1584
–7.
10
Sun J, Hedelin M, Zheng SL, et al. Interleukin-6 sequence variants are not associated with prostate cancer risk.
Cancer Epidemiol Biomarkers Prev
2004
;
13
:
1677
–9.
11
Wu HC, Chang CH, Chen HY, Tsai FJ, Tsai JJ, Chen WC. p53 gene codon 72 polymorphism but not tumor necrosis factor-alpha gene is associated with prostate cancer.
Urol Int
2004
;
73
:
41
–6.
12
McCarron SL, Edwards S, Evans PR, et al. Influence of cytokine gene polymorphisms on the development of prostate cancer.
Cancer Res
2002
;
62
:
3369
–72.
13
Gohagan JK, Prorok PC, Hayes RB, Kramer BS. The Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial of the National Cancer Institute: history, organization, and status.
Control Clin Trials
2000
;
21
:
251
–72S.
14
O'Brien B, Nichaman L, Browne JE, Levin DL, Prorok PC, Gohagan JK. Coordination and management of a large multicenter screening trial: the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial.
Control Clin Trials
2000
;
21
:
310
–28S.
15
Hayes RB, Reding D, Kopp W, et al. Etiologic and early marker studies in the prostate, lung, colorectal and ovarian (PLCO) cancer screening trial.
Control Clin Trials
2000
;
21
:
349
–55S.
16
Packer BR, Yeager M, Staats B, et al. SNP500Cancer: a public resource for sequence validation and assay development for genetic variation in candidate genes.
Nucleic Acids Res
2004
;
32
Database issue:
D528
–32.
17
Lewontin RC, Feldman MW. A general asymptotic property of two-locus selection models.
Theor Popul Biol
1988
;
34
:
177
–93.
18
Devlin B, Risch N. A comparison of linkage disequilibrium measures for fine-scale mapping.
Genomics
1995
;
29
:
311
–22.
19
Schaid DJ, Rowland CM, Tines DE, Jacobson RM, Poland GA. Score tests for association between traits and haplotypes when linkage phase is ambiguous.
Am J Hum Genet
2002
;
70
:
425
–34.
20
Lake SL, Lyon H, Tantisira K, et al. Estimation and tests of haplotype-environment interaction when linkage phase is ambiguous.
Hum Hered
2003
;
55
:
56
–65.
21
Giri D, Ozen M, Ittmann M. Interleukin-6 is an autocrine growth factor in human prostate cancer.
Am J Pathol
2001
;
159
:
2159
–65.
22
Smith PC, Hobisch A, Lin DL, Culig Z, Keller ET. Interleukin-6 and prostate cancer progression.
Cytokine Growth Factor Rev
2001
;
12
:
33
–40.
23
Koch AE, Polverini PJ, Kunkel SL, et al. Interleukin-8 as a macrophage-derived mediator of angiogenesis.
Science
1992
;
258
:
1798
–801.
24
Strieter RM, Kunkel SL, Elner VM, et al. Interleukin-8. A corneal factor that induces neovascularization.
Am J Pathol
1992
;
141
:
1279
–84.
25
Moore BB, Arenberg DA, Stoy K, et al. Distinct CXC chemokines mediate tumorigenicity of prostate cancer cells.
Am J Pathol
1999
;
154
:
1503
–12.
26
Inoue K, Slaton JW, Eve BY, et al. Interleukin 8 expression regulates tumorigenicity and metastases in androgen-independent prostate cancer.
Clin Cancer Res
2000
;
6
:
2104
–19.
27
Veltri RW, Miller MC, Zhao G, et al. Interleukin-8 serum levels in patients with benign prostatic hyperplasia and prostate cancer.
Urology
1999
;
53
:
139
–47.
28
Moore KW, de Waal Malefyt R, Coffman RL, O'Garra A. Interleukin-10 and the interleukin-10 receptor.
Annu Rev Immunol
2001
;
19
:
683
–765.
29
de Waal Malefyt R, Abrams J, Bennett B, Figdor CG, de Vries JE. Interleukin 10 (IL-10) inhibits cytokine synthesis by human monocytes: an autoregulatory role of IL-10 produced by monocytes.
J Exp Med
1991
;
174
:
1209
–20.
30
Stearns ME, Rhim J, Wang M. Interleukin 10 (IL-10) inhibition of primary human prostate cell-induced angiogenesis: IL-10 stimulation of tissue inhibitor of metalloproteinase-1 and inhibition of matrix metalloproteinase (MMP)-2/MMP-9 secretion.
Clin Cancer Res
1999
;
5
:
189
–96.