Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Feb 1.
Published in final edited form as: Cancer Causes Control. 2009 Nov 10;21(2):289–300. doi: 10.1007/s10552-009-9461-5

Genetic Variation in DNA Repair Genes and Prostate Cancer Risk: Results from a Population-Based Study

Ilir Agalliu 1, Erika M Kwon 2, Claudia A Salinas 3,4,*, Joseph S Koopmeiners 3,5, Elaine A Ostrander 2, Janet L Stanford 3,4
PMCID: PMC2811225  NIHMSID: NIHMS165625  PMID: 19902366

Abstract

Objective

DNA repair pathways are crucial to prevent accumulation of DNA damage and maintain genomic stability. Alterations of this pathway have been reported in many cancers. An increase in oxidative DNA damage or decrease of DNA repair capacity with aging or due to germline genetic variation may affect prostate cancer risk.

Methods

Pooled data from two population-based studies (1,457 cases and 1,351 controls) were analyzed to examine associations between 28 SNPs in 9 DNA repair genes (APEX1, BRCA2, ERCC2, ERCC4, MGMT, MUTYH, OGG1, XPC, and XRCC1) and prostate cancer risk. We also explored whether associations varied by smoking, by family history or clinical features of prostate cancer.

Results

There were no associations between these SNPs and overall risk of prostate cancer. Risks did not vary either by smoking or by family history of prostate cancer. Although, two SNPs in BRCA2 (rs144848, rs1801406) and two SNPs in ERCC2 (rs1799793, rs13181) showed stronger associations with high Gleason score or advanced stage tumors when comparing homozygous men carrying the minor vs. major allele, results were not statistically significantly different between clinically aggressive and non-aggressive tumors.

Conclusion

Overall this study found no associations between prostate cancer and the SNPs in DNA repair genes. Given the complexity of this pathway and its crucial role in maintenance of genomic stability a pathway-based analysis of all 150 genes in DNA repair pathways, as well as exploration of gene-environment interactions may be warranted.

Keywords: DNA repair, SNPs, prostate cancer, case-control study, genetic variation

Introduction

Prostate cancer is the most commonly diagnosed solid tumor and the second leading cause of cancer-related deaths among men in the U.S. (1). It is a multi-factorial disease, where both genetic and environmental factors contribute to disease incidence (2). The exponential increase in risk of prostate cancer associated with aging may reflect the accumulation of DNA damage as a result of a series of processes including oxidative stress, inflammation or environmental carcinogens or a decrease in DNA damage-repair response capacity (3–5). An increase in generation of reactive oxygen species from metabolism of endogenous (e.g., androgens, inflammation) and exogenous (e.g., dietary meat, fat, environmental toxins) compounds within the prostate cells can produce DNA adducts or directly damage DNA (4–7). In support of this hypothesis, several studies have reported that DNA adducts are formed in the prostate tissue as a result of exposure to oxidative stress or environmental toxins particularly heterocyclic amines (derived from red meat mutagens) and polycyclic aromatic hydrocarbons (8–10).

DNA repair mechanisms are important pathways in removal of oxidative DNA compounds or DNA adducts from damaged genomic sites (11). There are a number of DNA repair pathways, each responsible for repairing a different type of DNA damage. Base excision repair removes single base modifications including single strand breaks, oxidative DNA damage and non-bulky adducts, where as nucleotide excision repair removes larger lesions, which often result from environmental exposures such as smoking, UV radiation or external carcinogens (11, 12). Alkyltransferases directly reverse DNA damage by transferring alkyl groups from damaged DNA onto the transferase enzyme (11). Finally, double-strand DNA breaks are repaired through complex mechanisms including homologous recombination and end-joining repair pathways (13, 14). Germline genetic variation in the above DNA repair genes/pathway, which may affect the capacity of encoded DNA repair enzymes to effectively remove DNA adducts or lesions, may result in enhanced cancer risk (15–18).

Several epidemiologic studies have examined associations between single-nucleotide polymorphisms (SNPs) in DNA repair genes, mostly non-synonymous SNPs with potential functional significance, and risk of prostate cancer (17, 19–26). However, results have been inconsistent across these studies due in part to different study populations, case ascertainment, or due to small sample sizes of each study and thus the potential for false-positive findings as well as limited power to detect modest associations. The objective of our study was to examine associations between genetic variants in DNA repair genes and risk of prostate cancer in a large population-based case-control dataset, with detailed information on demographic and lifestyle factors, screening as well as clinical features of the disease. In addition, we were interested in exploring whether SNP-prostate cancer associations varied according to smoking status, family history of prostate cancer, or clinical features of this cancer.

Materials and Methods

Study Population

Study subjects were Caucasian and African American men residing in King County, Washington, who participated in one of two population-based case-control studies of prostate cancer with ascertainment periods from 1993 to 1996 and 2002 to 2005 that have been previously described (27, 28). Incident cases with histologically confirmed adeno-carcinoma of the prostate were 35 to 74 years old at diagnosis, and were identified via the Seattle-Puget Sound Surveillance, Epidemiology, and End Results (SEER) cancer registry. The SEER cancer registry also provided information on clinical characteristics of prostate cancer including Gleason score, tumor stage, and serum prostate-specific antigen (PSA) levels at diagnosis, as well as primary treatment for prostate cancer. Controls were men without a self-reported physician’s diagnosis of prostate cancer, identified via random digit dialing, frequency matched to cases by 5-year age groups, and recruited evenly throughout the ascertainment period of the cases. Both studies were approved by the Institutional Review Board (IRB) of the Fred Hutchinson Cancer Research Center, and written informed consent was obtained from all study participants. Genotyping was approved by the IRB of the National Human Genome Research Institute.

Study subjects completed a structured in-person interview administered by trained male interviewers. The questionnaire collected information about demographic, social and lifestyle characteristics, medical history including prostate cancer screening, as well as information about family history of prostate cancer. After the interview, participants were asked to provide a blood sample. Of the combined 2,244 eligible prostate cancer cases and 2,448 eligible controls identified in the two population-based studies, 1,754 (78.2%) cases and 1,645 (67.2%) controls agreed to participate and were interviewed. Among men who participated, 1,457 (83.1%) cases and 1,351 (82.1%) controls provided a blood sample, which was used to obtain germline DNA for genotyping.

Selection of SNPs and Genotyping

SNPs in DNA repair genes were selected for genotyping in this study based on the following criteria: (a) functional significance, as assessed by potential effects on enzyme level activity or projected influence on DNA adducts levels, (b) previously reported association(s) with prostate cancer risk in earlier studies, (c) tagSNPs with a minor allele frequency ≥5% selected from available HapMap (version 3.1), dbSNP (http://www.ncbi.nlm.nih.gov/projects/SNP/) and Genome Variation Server (http://pga.gs.washington.edu/gvs). The majority of the SNPs in 9 DNA repair genes genotyped in our study were non-synonomous SNP (n=13) followed by synonomous SNPs (n=8), SNPs located in the UTR region (n=4) and intronic SNPs (n=4). Of the 28 SNPs successfully genotyped, 21 were tagging SNP; however, the coverage of variation across large genes such as BRCA2 or MGMT is small. SNPs with prior evidence from association analyses reported from other studies were included, except for the SNPs in the mismatch DNA repair pathway, which were part of another analysis. Potential functional prediction of SNPs that resulted in amino-acid changes was assessed using in silico classification program Polyphen (29).

Genomic DNA was purified from peripheral lymphocytes using standard protocols (30). The Applied Biosystems (ABI) SNPlex™ Genotyping system was used for genotyping and GeneMapper software was used for calling alleles. Discrimination of the specific SNP allele was carried out with the Applied Biosystems 3730xl DNA analyzer and was based on the presence of a unique sequence assigned to the original allele-specific oligonucleotide. Quality control included genotyping of 140 blind duplicate samples that were distributed across all genotyping batches. Each batch of DNA aliquots that were genotyped incorporated similar numbers of cases and controls samples collected in a specific calendar year (e.g. 2003) and samples were distributed equally across the entire genotyping plates. The laboratory personnel that performed genotyping were however blinded to the case-control status of the samples.

Initially 35 SNPs in 10 DNA repair genes were selected. Of these, 4 SNPs (APEX1 rs2307486, MUTYH rs3219489, XRCC1 rs3547, XRCC3 rs861539) failed on the genotyping platform, and two others (ERCC4 rs1799802, MGMT rs2308318) were monomorphic in our study population. One SNP (rs1052133) in the OGG1 had a drop-out rate of 9.5%, and thus was excluded from all statistical analyses. The remaining 28 SNPs in 9 DNA repair genes with an average drop-out of 0.6% and an average genotype concordance of 99.7% across 140 blind duplicate samples were included in statistical analyses.

Statistical Analyses

Deviation of genotype frequencies from Hardy-Weinberg Equilibrium (HWE) among Caucasian and African-American controls considered separately was assessed by χ2-tests (31). Unconditional logistic regression was used to examine associations between SNP genotypes and prostate cancer risk among Caucasian and African-American men and to compute odds ratios (OR) and 95% confidence intervals (CI) (32). Confounding by established and potential risk factors for prostate cancer was assessed for each genotype separately, fitting models using each main effect and then evaluating the change in parameter estimates of the SNP genotypes when other variables entered the models one at a time. All analyses were adjusted for age at diagnosis for cases and at reference date for controls. Further adjustment for a first-degree family history of prostate cancer, history of screening for prostate cancer (PSA and DRE tests) and smoking did not change the risk estimates and thus were not included in the final models. Both additive and dominant genetic models were considered for each variant allele depending on the distribution of genotypes. Likelihood ratio-based test statistics were used to identify statistically significant associations between SNP genotypes and prostate cancer risk.

A permutation procedure was used to account for the effect of multiple testing. Pairs of case-control labels and ages were permuted in order to approximate the distribution of the age-adjusted p-values under the null hypothesis. For each permutation codominant and dominant models were fit for all SNPs and the minimum of these p-values were kept for each SNP. The p-values were then ordered to approximate the null distribution of the order statistics for the p-values, starting from the smallest to the largest. The original p-values were also ordered and permutation p-values were calculated by comparing the ordered p-values to the null distribution for the appropriate order statistic. Permutation p-values can be interpreted as the probability of observing a p-value less than or equal to what was observed for the given order statistic under the null hypothesis of no association with disease for any of the 28 SNPs. The permutation approach to approximating the null distribution of the order statistics will be valid regardless of any correlation between the SNPs. A SNP was considered to be significantly associated with prostate cancer risk if the nominal p-value and the permuted p-value were both <0.05.

In addition, interactions between SNP genotypes and first-degree family history of prostate cancer (yes vs. no) or between base- or nucleotide excision repair genes and either smoking status or lifetime pack-years of smoking were examined in relation to prostate cancer risk. Interactions were tested by including an interaction term in the logistic regression models and comparing the fully saturated model containing the main effects and interaction term with the reduced model containing only the main effects using a likelihood ratio test to determine statistical significance of the interaction effects (33). Lastly, we also examined associations between SNP genotypes and clinical characteristics of prostate cancer. With respect to the Gleason score analyses, cases were grouped into those with Gleason scores of 2–6 or 7=3+4, and those with Gleason scores of 7=4+3 or 8–10 at diagnosis. For cancer stage, cases diagnosed with regional or distant stage were compared to men with localized stage at diagnosis. The frequency of genotypes for DNA repair SNPs in each group of cases was compared to the frequency of genotypes in controls using polytomous logistic regression models (34); these final models were adjusted for age and prostate cancer screening history.

Results

Selected characteristics of prostate cancer cases and controls, stratified by race, are presented in Table 1. The distribution of age was similar between Caucasian cases and controls; however, among African Americans cases were slightly older than controls (p=0.0002) although this could have been due to small number of cases and controls in this category. In both racial groups, cases were more likely than controls to report a first-degree family history of prostate cancer and to have undergone PSA screening within the 5 year-period before diagnosis or reference date. Cases and controls were similar with respect to other factors including body mass index, education and smoking. With respect to clinical features, the majority of prostate cancer cases were diagnosed with localized stage or Gleason score 2–6 or 7(3+4) cancers and the distribution of Gleason score and tumor stage was similar between cases ascertained in two different periods: 1993–1996 and 2002–2005 (data not shown).

Table 1.

Selected characteristics of prostate cancer cases and controls genotyped for SNPs in DNA repair genes, by race

Caucasians African - Americans
Characteristic Cases (N=1,308) Controls (N=1,266) P-value† Cases (N=149) Controls (N=85) P-value†
Age (years); n (%) 0.36 0.0002
 35–49 102 (7.8) 107 (8.5) 16 (10.7) 19 (22.4)
 50–54 188 (14.4) 178 (14.1) 26 (17.5) 31 (36.5)
 55–59 325 (24.9) 343 (27.1) 32 (21.5) 15 (17.7)
 60–64 395 (30.2) 334 (26.3) 38 (25.5) 14 (16.5)
 65–69 153 (11.7) 160 (12.6) 24 (16.1) 4 (4.7)
 70–74 145 (11.1) 144 (11.4) 13 (8.7) 2 (2.4)
First-degree family history of prostate cancer; n (%) <0.0001 0.04
 No 1,025 (78.4) 1,125 (88.9) 119 (79.9) 75 (88.2)
 Yes 283 (21.6) 141 (11.1) 30 (20.1) 10 (11.8)
Body mass index (BMI), kg/m2 n (%) 0.34 0.39
 < 25.0 429 (32.8) 389 (30.7) 38 (25.5) 21 (24.7)
 25.0–29.9 637 (48.7) 618 (48.8) 69 (46.1) 33 (38.8)
 ≥ 30 242 (18.5) 259 (20.5) 42 (28.2) 31 (36.5)
Smoking status; n (%) 0.13 0.64
 Non smoker 522 (39.9) 541 (42.7) 63 (42.3) 31 (36.5)
 Former smoker 631 (48.2) 561 (44.3) 45 (30.2) 30 (35.3)
 Current smoker 155 (11.9) 164 (13.0) 41 (27.5) 24 (28.2)
Education; n (%) 0.85 0.09
 High school or less 230 (17.6) 237 (18.7) 57 (38.3) 24 (28.2)
 Some college 324 (24.8) 300 (23.7) 49 (32.9) 23 (27.1)
 College degree 369 (28.2) 354 (28.0) 21 (14.1) 21 (24.7)
 Graduate degree 385 (29.4) 375 (29.6) 22 (14.7) 17 (20.0)
Prostate cancer screening history‡; n (%) <0.0001 0.002
 None 135 (10.3) 168 (13.3) 22 (14.8) 15 (17.7)
 DRE only 223 (17.1) 482 (38.1) 34 (22.8) 36 (42.3)
 PSA 950 (72.6) 616 (48.7) 93 (62.4) 34 (40.0)
Gleason score, n (%)
 2–4 66 (5.1) 5 (3.4)
 5–6 681 (52.1) 61 (40.9)
 7 (3+4) 355 (27.1) 53 (35.6)
 7 (4+3) 76 (5.8) 14 (9.4)
 8–10 126 (9.6) 14 (9.4)
 Missing 4 (0.3) 2 (1.3)
Stage of cancer, n (%)
 Localized 1,022 (78.1) 118 (79.2)
 Regional 254 (19.4) 26 (17.5)
 Distant 32 (2.5) 5 (3.4)
PSA valueξ (ng/ml); n (%) <0.0001 <0.0001
 0.01–3.9 178 (13.6) 1,175 (92.8) 11 (7.4) 78 (91.8)
 4.0–9.9 705 (53.9) 74 (5.9) 91 (61.1) 6 (7.1)
 10.0–19.9 188 (14.4) 15 (1.2) 19 (12.8) 1 (1.2)
 ≥ 20 118 (9.0) 2 (0.2) 20 (13.4) -
 Missing 119 (9.1) - 8 (5.4) -
Open in a new tab
†

Chi-square p-value;

‡

Prostate cancer screening history within the 5 years before prostate cancer diagnosis or reference date.

ξ

Serum PSA value at prostate cancer diagnosis for cases and at interview for controls Abbreviations: DRE - digital rectal examination; PSA - prostate specific antigen.

All 28 SNPs evaluated in this study were in HWE among both Caucasian and African Americans controls (all p>0.05). Table 2 presents associations between SNP genotypes and risk of prostate cancer in Caucasians and African-Americans, separately. In single SNP analyses, modest associations with overall risk of prostate cancer were observed in Caucasians for BRCA2 rs1801406 (OR=0.81; 95% CI 0.69–0.95; comparing any G allele vs. homozygous A allele carriers) and ERCC2 rs1799793 (OR=0.70; 95% CI 0.54–0.91; comparing men with AA vs. GG genotype). In African-Americans there was an association between prostate cancer and BRCA2 rs543304 where men with any C allele had an OR=0.54 (95% CI 0.30–0.99) in comparison to men with the TT genotype. However, after adjusting for multiple comparisons the permuted P-values of associations between these three SNPs and risk of prostate cancer were no longer statistically significant.

Table 2.

Associations between SNPs in DNA repair genes and prostate cancer risk, by race

Caucasians African-American
Cases Controls Cases Controls
Gene SNP Genotype N % N % OR† 95% CI Permuted P* N % N % OR† 95% CI Permuted P*
APEX1 rs1320150 AA 424 33.5 415 33.4 1.00 - 0.42 57 40.1 37 46.8 1.00 - 0.84
AG 612 48.4 615 49.5 0.97 0.82–1.16 65 45.8 33 41.8 1.43 0.77–2.66
GG 229 18.1 212 17.1 1.06 0.84–1.33 20 14.1 9 11.4 1.48 0.58–3.79
AG + GG 841 66.5 827 66.6 1.00 0.84–1.18 85 59.9 42 53.2 1.44 0.80–2.59
APEX1 rs2275007 GG 420 33.4 415 33.5 1.00 - 0.79 61 43.0 37 46.8 1.00 - 0.53
AG 608 48.3 614 49.5 0.98 0.82–1.17 62 43.7 31 39.2 1.40 0.75–2.62
AA 230 18.3 211 17.0 1.08 0.85–1.36 19 13.4 11 13.9 1.08 0.44–2.65
AG + AA 838 66.6 825 66.5 1.01 0.85–1.19 81 57.0 42 53.2 1.32 0.73–2.36
BRCA2 rs144848 TT 655 51.6 654 52.6 1.00 - 0.52 104 73.2 59 74.7 1.00 - 0.36
GT 498 39.4 500 40.2 0.99 0.84–1.17 36 25.4 18 22.8 0.98 0.49–1.95
GG 116 9.1 89 7.2 1.30 0.98–1.75 2 1.4 2 2.5 0.48 0.06–3.85
GT + GG 614 48.5 589 47.4 1.04 0.89–1.22 38 26.8 20 25.3 0.93 0.48–1.81
BRCA2 rs15869 AA 770 60.8 787 63.3 1.00 - 0.59 136 95.1 71 89.9 1.00 - 0.90
AC 439 34.7 398 32.0 1.13 0.95–1.33 7 4.9 7 8.9 0.62 0.20–1.91
CC 57 4.5 59 4.7 0.99 0.68–1.44 - - 1 1.3 - -
AC + CC 496 39.2 457 36.7 1.11 0.94–1.30 7 4.9 8 10.1 0.53 0.18–1.58
BRCA2 rs1801406 AA 635 50.5 563 45.2 1.00 - 0.22 79 54.5 46 58.3 1.00 - 0.53
AG 501 39.9 556 44.6 0.80 0.68–0.94 58 40.0 31 39.2 1.13 0.62–2.06
GG 121 9.6 127 10.2 0.84 0.64–1.11 8 5.5 2 2.5 3.34 0.65–17.12
AG + GG 622 49.5 683 54.8 0.81 0.69–0.95 66 45.5 33 41.7 1.25 0.70–2.25
BRCA2 rs543304 TT 821 65.1 839 67.6 1.00 - 0.65 98 68.5 44 55.7 1.00 - 0.72
CT 399 31.6 364 29.3 1.12 0.94–1.33 42 29.4 30 38.0 0.59 0.32–1.10
CC 41 3.3 38 3.1 1.11 0.70–1.74 3 2.1 5 6.3 0.27 0.06–1.21
CT + CC 440 34.9 402 32.4 1.12 0.95–1.32 45 31.5 35 44.3 0.54 0.30–0.99
ERCC2 (XPD) rs1052555 CC 591 46.9 565 45.4 1.00 - 0.47 110 76.4 60 76.0 1.00 - 0.31
CT 549 43.6 535 43.0 0.98 0.83–1.16 29 20.1 17 21.5 0.82 0.40–1.68
TT 119 9.5 144 11.6 0.79 0.60–1.03 5 3.5 2 2.5 1.56 0.27–8.86
CT + TT 668 53.1 679 54.6 0.94 0.80–1.10 34 23.6 19 24.0 0.90 0.45–1.76
ERCC2 (XPD) rs13181 TT 505 41.0 480 39.1 1.00 - 0.17 87 59.6 50 60.2 1.00 - 0.41
GT 575 46.6 571 46.5 0.96 0.81–1.13 48 32.9 28 33.7 0.85 0.46–1.57
GG 153 12.4 177 14.4 0.82 0.64–1.05 11 7.5 5 6.0 1.19 0.37–3.78
GT + GG 728 59.0 748 60.9 0.92 0.79–1.09 59 40.4 33 39.7 0.90 0.50–1.61
ERCC2 (XPD) rs1799793 GG 545 44.0 527 43.2 1.00 - 0.08 106 73.6 65 79.3 1.00 - 0.77
AG 575 46.4 528 43.2 1.05 0.89–1.25 31 21.5 15 18.3 1.28 0.62–2.64
AA 120 9.7 166 13.6 0.70 0.54–0.91 7 4.9 2 2.4 2.35 0.44–12.43
AG + AA 695 56.1 694 56.8 0.97 0.83–1.13 38 26.4 17 20.7 1.40 0.71–2.77
ERCC2 (XPD) rs238406 GG 365 29.0 383 30.9 1.00 - 0.24 112 77.8 65 80.2 1.00 - 0.14
GT 636 50.4 600 48.5 1.11 0.93–1.33 29 20.1 16 19.8 0.98 0.48–2.00
TT 260 20.6 255 20.6 1.07 0.86–1.34 3 2.1 0 0.0 - -
GT + TT 896 71.1 855 69.1 1.10 0.93–1.31 32 22.2 16 19.8 1.09 0.54–2.20
ERCC4 (XPF) rs1799801 TT 633 50.3 623 50.1 1.00 - 0.91 96 66.2 55 69.6 1.00 - 0.72
CT 521 41.4 521 41.9 0.98 0.84–1.16 46 31.7 23 29.1 1.27 0.67–2.40
CC 105 8.3 99 8.0 1.04 0.78–1.40 3 2.1 1 1.3 3.02 0.29–30.9
CT + CC 626 49.7 620 49.9 0.99 0.85–1.16 49 33.8 24 30.4 1.33 0.71–2.49
ERCC4 (XPF) rs1800067 GG 1025 84.0 1012 83.0 1.00 - 0.79 136 94.4 78 96.3 1.00 - 0.59
AG 183 15.0 202 16.6 0.89 0.72–1.11 8 5.6 3 3.7 2.12 0.52–8.54
AA 13 1.1 5 0.4 2.58 0.92–7.25 - - - - - -
AG + AA 196 16.1 207 17.0 0.93 0.75–1.16
MGMT rs12917 CC 949 75.9 916 74.1 1.00 - 0.87 106 72.1 60 74.1 1.00 - 0.68
CT 269 21.5 298 24.1 0.87 0.72–1.05 35 23.8 20 24.7 1.18 0.61–2.29
TT 32 2.6 23 1.9 1.34 0.78–2.31 6 4.1 1 1.2 2.75 0.30–25.41
CT + TT 301 24.1 321 26.0 0.91 0.75–1.09 41 27.9 21 25.9 1.26 0.66–2.40
MGMT rs2308321 AA 926 76.7 922 77.2 1.00 - 0.50 130 92.9 73 91.3 1.00 - 0.39
AG 267 22.1 256 21.4 1.04 0.86–1.26 10 7.1 7 8.8 0.82 0.29–2.32
GG 14 1.2 17 1.4 0.82 0.40–1.67 - - - - - -
AG + GG 281 23.3 273 22.8 1.02 0.85–1.24
MGMT rs2308327 AA 950 76.1 960 77.1 1.00 - 0.75 134 92.4 75 90.4 1.00 - 0.31
AG 276 22.1 266 21.4 1.05 0.87–1.27 11 7.6 8 9.6 0.74 0.27–2.00
GG 22 1.8 20 1.6 1.11 0.60–2.05 - - - - - -
AG + GG 298 23.9 286 23.0 1.05 0.87–1.26
MUTYH rs3219484 GG 1085 86.4 1070 85.7 1.00 - 0.77 140 97.9 75 94.9 1.00 - 0.77
AG 164 13.1 173 13.9 0.94 0.74–1.18 3 2.1 4 5.1 0.29 0.06–1.36
AA 7 0.6 5 0.4 1.39 0.44–4.41 - - - - - -
AG + AA 171 13.6 178 14.3 0.95 0.76–1.19
MUTYH rs9429072 AA 720 57.1 686 55.0 1.00 - 0.37 34 23.5 24 30.4 1.00 - 0.54
AG 455 36.1 485 38.9 0.89 0.76–1.05 71 49.0 32 40.5 1.39 0.69–2.79
GG 85 6.8 76 6.1 1.07 0.77–1.48 40 27.6 23 29.1 1.11 0.51–2.39
AG + GG 540 42.9 561 45.0 0.92 0.78–1.07 111 76.6 55 69.6 1.27 0.67–2.42
OGG1 rs3218997 CC 1126 91.4 1137 91.5 1.00 - 0.84 139 95.9 77 97.5 1.00 - 0.52
CT 106 8.6 106 8.5 1.01 0.76–1.34 6 4.1 2 2.5 1.82 0.34–9.78
XPC rs1126547 CC 969 76.8 938 75.4 1.00 - 0.33 109 75.7 61 77.2 1.00 - 0.53
CG 273 21.6 287 23.1 0.92 0.76–1.11 30 20.8 17 21.5 0.93 0.46–1.89
GG 20 1.6 19 1.5 1.01 0.54–1.91 5 3.5 1 1.3 1.92 0.21–17.34
CG + GG 293 23.2 306 24.6 0.93 0.77–1.11 35 24.3 18 22.8 0.99 0.50–1.97
XPC rs2228001 AA 457 36.4 461 36.9 1.00 - 0.66 70 47.6 36 43.4 1.00 - 0.23
AC 595 47.3 600 48.0 1.00 0.84–1.19 61 41.5 38 45.8 0.80 0.44–1.46
CC 205 16.3 190 15.2 1.09 0.86–1.38 16 10.9 9 10.8 1.04 0.40–2.69
AC + CC 800 63.6 790 63.2 1.02 0.87–1.20 77 52.4 47 56.6 0.85 0.48–1.49
XPC rs2733532 CC 461 36.5 467 37.6 1.00 - 0.74 72 50.0 39 47.6 1.00 - 0.41
CT 602 47.7 599 48.2 1.02 0.86–1.21 57 39.6 35 42.7 0.91 0.50–1.65
TT 199 15.8 176 14.2 1.14 0.90–1.46 15 10.4 8 9.8 1.19 0.45–3.20
CT + TT 801 63.5 775 62.4 1.05 0.89–1.23 72 50.0 43 52.4 0.96 0.54–1.69
XPC rs2733534 GG 340 27.6 311 25.4 1.00 - 0.39 58 40.0 39 47.6 1.00 - 0.66
CG 607 49.2 635 51.8 0.88 0.72–1.06 64 44.1 35 42.7 1.24 0.67–2.28
CC 286 23.2 279 22.8 0.94 0.75–1.18 23 15.9 8 9.8 1.88 0.74–4.79
CG + CC 893 72.4 914 74.6 0.90 0.75–1.07 87 60.0 43 52.4 1.36 0.77–2.42
XPC rs8516 TT 771 61.1 753 60.3 1.00 - 0.85 106 73.6 66 83.5 1.00 - 0.76
CT 425 33.7 436 34.9 0.95 0.81–1.13 38 26.4 13 16.5 1.73 0.83–3.60
CC 66 5.2 60 4.8 1.08 0.75–1.55 - - - - - -
CT + CC 491 38.9 496 39.7 0.97 0.83–1.14
XRCC1 rs1799782 CC 1098 88.1 1071 86.7 1.00 - 0.23 131 89.7 72 86.8 1.00 - 0.60
CT 143 11.5 158 12.8 0.88 0.69–1.12 15 10.3 9 10.8 0.97 0.38–2.46
TT 5 0.4 6 0.5 0.82 0.25–2.68 - - 2 2.4 - -
CT + TT 148 11.9 164 13.3 0.88 0.69–1.11 15 10.3 11 13.2 0.77 0.32–1.86
XRCC1 rs25487 GG 522 41.5 481 38.8 1.00 - 0.77 103 71.5 53 64.6 1.00 - 0.83
AG 576 45.8 590 47.6 0.90 0.76–1.07 37 25.7 27 33.0 0.67 0.36–1.25
AA 159 12.7 169 13.6 0.87 0.68–1.11 4 2.8 2 2.4 0.97 0.16–5.84
AG + AA 735 58.5 759 61.2 0.89 0.76–1.05 41 28.5 29 35.4 0.69 0.38–1.27
XRCC1 rs25489 GG 1120 90.0 1145 91.4 1.00 - 0.26 137 93.8 76 91.6 1.00 - 0.61
AG 121 9.7 106 8.5 1.17 0.87–1.53 9 6.2 7 8.4 0.59 0.20–1.76
AA 3 0.2 2 0.2 1.55 0.26–9.27 - - - - - -
AG + AA 124 10.0 108 8.6 1.17 0.89–1.54
XRCC1 rs2854509 CC 798 63.4 777 62.8 1.00 - 0.66 98 67.6 56 70.0 1.00 - 0.21
AC 406 32.3 400 32.3 0.99 0.83–1.17 39 26.9 23 28.8 1.09 0.57–2.08
AA 55 4.4 61 4.9 0.88 0.60–1.28 8 5.5 1 1.3 2.94 0.34–25.66
AC + AA 461 36.6 461 37.2 0.97 0.83–1.14 47 32.4 24 30.0 1.19 0.64–2.21
XRCC1 rs915927 AA 400 31.8 409 32.8 1.00 - 0.73 62 42.8 30 38.0 1.00 - 0.41
AG 622 49.4 618 49.6 1.03 0.86–1.23 54 37.2 38 48.1 0.79 0.42–1.50
GG 238 18.9 220 17.6 1.11 0.88–1.39 29 20.0 11 13.9 1.32 0.56–3.11
AG + GG 860 68.3 838 67.2 1.05 0.89–1.24 83 57.2 49 62.0 0.92 0.51–1.65
Open in a new tab

Total number of cases and controls in both Caucasians and African-Americans vary by SNPs due to missing genotype data.

†

ORs and 95% CI are adjusted for age at diagnosis (cases) and age at reference date (controls).

*

Permuted P-values are for the co-dominant models (2 degrees of freedom) and present the adjusted P-values for multiple comparisons using a permutation procedure (see statistical methods for details).

Next we explored interactions between DNA repair SNPs and first-degree family history of prostate cancer (yes vs. no) and smoking (both smoking status and lifetime pack-years of smoking) in relation to prostate cancer risk. There was no evidence for effect modification (data not shown). Lastly, we examined associations between DNA repair SNPs and clinical characteristics of prostate cancer including Gleason score and tumor stage among Caucasians (Table 3). With respect to clinical features of this disease two SNPs (rs144848, rs1801406) in BRCA2 and two SNP (rs1799793, rs13181) in ERCC2 showed some associations with Gleason score or tumor stage in single SNP analyses. For BRCA2 rs144848, although men with the GG genotype had an OR of 1.83 (95% CI 1.09–3.08) of high Gleason score tumors in comparison to men with the TT genotype, the risk estimate was not statistically significantly different in comparison to the OR obtained for tumors with Gleason score 2–6 or 7(3+4) (OR=1.32, p-value for differences in ORs estimates=0.22). For BRCA2 rs1801406 and ERCC2 rs1799793, although men homozygous for the minor allele had a slightly stronger reduction in risk of high Gleason score tumors in comparison to men homozygous for the major allele (ORs of 0.48 and 0.54, respectively) the risk estimates between higher and lower Gleason score tumors were not statistically significantly different. Similar findings of no significant differences were also observed in relation to tumor stage when data were stratified according to localized tumors versus regional/distant tumors. No associations were observed between the remaining 24 SNPs in DNA repair genes and any of the clinical characteristics of prostate cancer.

Table 3.

Associations between SNPs in DNA repair genes and clinical features of prostate cancer among Caucasians

Gleason Score Controls Cases*: Gleason score 2–6 or 7 (3+4) Cases*: Gleason score 7 (4+3) or 8–10 p-value OR1≠ OR2
Gene SNP Genotype N % N % OR1† 95% CI N % OR2† 95% CI
BRCA2 rs144848 TT 654 52.6 562 52.5 1.00 - 92 47.2 1.00 -
GT 500 40.2 416 38.9 0.99 0.83–1.18 81 41.5 1.16 0.84–1.60 0.34
GG 89 7.2 92 8.6 1.32 0.95–1.83 22 11.3 1.83 1.09–3.08 0.22
BRCA2 rs1801406 AA 563 45.2 534 50.4 1.00 - 98 50.5 1.00 -
AG 556 44.6 415 39.2 0.75 0.63–0.90 85 43.8 0.86 0.63–1.18 0.42
GG 127 10.2 110 10.4 0.87 0.65–1.17 11 5.7 0.48 0.25–0.93 0.08
ERCC2 rs13181 TT 480 39.1 418 40.4 1.00 - 84 43.5 1.00 -
GT 571 46.5 492 47.5 0.96 0.80–1.15 82 42.5 0.80 0.57–1.11 0.28
GG 177 14.4 126 12.2 0.75 0.57–0.99 27 14.0 0.80 0.50–1.28 0.81
ERCC2 rs1799793 GG 527 43.2 450 43.1 1.00 - 92 47.7 1.00 -
AG 528 43.2 490 47.0 1.06 0.88–1.27 84 43.5 0.88 0.64–1.21 0.26
AA 166 13.6 103 9.9 0.67 0.50–0.89 17 8.8 0.54 0.31–0.93 0.45
Tumor Stage Controls Cases*: Localized Stage Cases*: Regional/Distant Stage p-value OR1 ≠ OR2
Gene SNP Genotype N % N % OR1† 95% CI N % OR2† 95% CI
BRCA2 rs144848 TT 654 52.6 512 51.6 1.00 - 143 51.8 1.00 -
GT 500 40.2 393 39.6 1.02 0.85–1.23 105 38.0 1.00 0.75–1.32 0.85
GG 89 7.2 88 8.9 1.37 0.99–1.90 28 10.1 1.57 0.98–2.51 0.57
BRCA2 rs1801406 AA 563 45.2 495 50.5 1.00 - 140 50.5 1.00 -
AG 556 44.6 389 39.7 0.76 0.64–0.92 112 40.4 0.79 0.60–1.04 0.84
GG 127 10.2 96 9.8 0.81 0.60–1.10 25 9.1 0.79 0.49–1.26 0.89
ERCC2 rs13181 TT 480 39.1 389 40.4 1.00 - 116 43.1 1.00 -
GT 571 46.5 450 46.7 0.94 0.78–1.14 125 46.5 0.89 0.67–1.19 0.74
GG 177 14.4 125 13.0 0.79 0.60–1.04 28 10.4 0.63 0.40–0.99 0.33
ERCC2 rs1799793 GG 527 43.2 422 43.5 1.00 - 123 45.6 1.00 -
AG 528 43.2 448 46.2 1.02 0.85–1.23 127 47.0 1.03 0.78–1.36 0.96
AA 166 13.6 100 10.3 0.68 0.51–0.91 20 7.4 0.50 0.30–0.83 0.24
Open in a new tab
*

Total number of cases and controls vary by SNPs due to missing genotype data. Cases with missing data on Gleason score (n=6) were excluded from respective analyses.

†

ORs and 95% CI are adjusted for age and prostate cancer screening history within the 5 years before prostate cancer diagnosis or reference date.

Discussion

In this population-based study we examined associations between 28 SNPs in 9 DNA repair genes and prostate cancer risk among 1,457 cases and 1,351 controls pooled from two prior studies conducted in King County, Washington. No associations were observed between any of these SNPs and overall risk of prostate cancer, after adjusting for age and multiple comparisons. Risks were not different according to family history of prostate cancer or by smoking (either smoking status or pack-years of smoking) with mostly null associations. With respect to clinical characteristics of prostate cancer, two SNPs in BRCA2 (rs144848, rs1801406) and two SNP in ERCC2 (rs1799793, rs13181) showed some associations with Gleason score and tumor stage in single SNP analysis; however ORs were not statistically significantly different between lower and higher Gleason score tumors or between localized versus regional or distant stage tumors.

In relation to BRCA2 SNPs, although both rs144848 (Asn372His) and rs1801406 (Lys1132 Lys) showed a stronger association with risk of higher grade prostate cancer in Caucasians, the ORs estimates were not statistically significant different from those obtained for Gleason 2–6 or 7(3+4) tumors. However, none of the above SNPs in BRCA2 were associated with tumor stage in our dataset. The BRCA2 is considered a strong susceptibility gene in prostate cancer, since studies of families segregating BRCA2 protein-truncating mutations (35–40), studies of populations who harbor founder mutations such as Icelandic (41) or Ashkenazi Jews (42–44), as well as studies of younger-onset prostate cancer (age at diagnosis < 55 yrs) unselected for family history (45, 46) all have reported that men who carry protein-truncating BRCA2 mutations have increased risk of prostate cancer with relative risk (RRs) ranging from 2.0 to 23.0 depending on several factors including study population, case ascertainment, age at prostate cancer diagnosis and familial history of several cancer such as prostate, breast and ovary. Unlike SNPs, which have an unclear role in protein function, the above deleterious mutations in BRCA2 result in an earlier truncation of the BRCA2 protein and thus are more likely to be linked to prostate cancer susceptibility; however given their very low prevalence in the general population (~0.1%) it is estimated that <1% of sporadic prostate cancers can be attributed to these disease-associated BRCA2 mutations (47).

With respect to ERCC2 (XPD) SNPs, we found no association with overall risk of prostate cancer; however there was an inverse association between two SNPs: rs1799793 (Asp312Asn) and rs13181 (Lys751Glu) and clinical features of prostate cancer including Gleason score and tumor stage. For both of these two non-synonymous SNPs although men homozygous for the minor allele had a stronger reduction in risk of higher Gleason score or regional or distant tumor stage in comparison to men with the more common genotype, these ORs were not statistically significant for those obtained for low Gleason score (2–6 or 7=3+4) and localized stage prostate cancer. The ERCC2 gene encodes for a protein that is part of the TFIIH complex, which unwinds the DNA helix around the lesion in the earlier steps of the nucleotide-excision repair (NER) pathway (12). This pathway is responsible for removal of DNA bulky lesions that occur from a variety of exposures, including UV-induced photoproducts, cross-links, oxidative damage or chemical adducts from PAH exposures (48) (11). Two other studies have examined associations between NER pathway and prostate cancer risk (17, 21). Hu and colleagues (17) examined associations between NER capacity in isolated prostate tissue as measured in a host-cell reaction assay, and prostate cancer risk is a small clinic-based study of 140 prostate cancer patients and 96 controls. They reported a 2.6-fold increased risk (95%CI 1.2–6.0) of prostate cancer for men in the lowest quartile of NER capacity in comparison to those in the highest quartile (17), however, they did not consider genetic variants in NER pathway genes including ERCC2. In another study, Rybicki and colleagues (21) examined associations between the same two variants in ERCC2 (rs1799793, rs13181) that we evaluated and risk of prostate cancer in a family-based study of 637 cases and 480 brother controls (the study population was primarily Caucasian) and reported a positive association between rs1799793 (Asp312Asn) and prostate cancer, where men with the AA genotype had a 60% increased risk of prostate cancer in comparison to men with the GG or GA genotype (21). No association was observed for ERCC2 rs13181 in that study (21), nor risks differed by clinical characteristics of prostate cancer We did not observe any association between these two SNPs (rs1799793, rs13181) and risk of prostate cancer in our study; however differences in results could be due several factors including false-positive findings or different study designs i.e. population-based vs. family-based (sibling brothers).

With respect to other SNPs in DNA repair genes we did not find any associations with overall risk of prostate cancer or clinical characteristics of this disease in our study population, although prior studies have reported positive associations between risk of prostate cancer and genetic variants in OGG1 (19), XRCC1 (20–24), MGMT (22) and XPC (25) in different populations. In a small study of 245 cases and 222 controls, Xu and colleagues (19) reported an OR of 3.2 (95% CI 1.19–8.73) for sporadic prostate cancer risk associated with the GG genotype vs. CC genotype for OGG1 rs1052133 (Ser326Cys). However that SNP was excluded from our analysis due to high frequency of drop-outs. In relation to XRCC1, two studies of Caucasian populations (20, 21) reported no associations between prostate cancer and rs25487 (Arg399Gln), rs1799782 (Arg194Trp) or rs25489 (Arg280His), which is similar to our findings. However, one of these studies reported effect modification between XRCC1 rs25487 (Arg399Gln) and ERCC2 rs1799793 (Asp312Asn) in relation to risk of prostate cancer (21). Finally, two small scale studies in Asian populations with a maximum of 250 cases, reported positive associations between risk of prostate cancer and MGMT Leu84Phe (22), XPC Lys939Gln (25) as well as XRCC1 Arg194Trp (22)(25). However the allele frequencies of these SNPs in Asians were different from those observed in our study population of Caucasians.

Our study has several strengths and limitations that should be carefully considered when interpreting the results. Strengths of the current study include the population-based design, the sample size, and the availability of information on potential confounding variables, as well as the availability of clinical data on prostate cancer cases. A limitation of our study is that we examined only a small number of SNPs in DNA repair genes with respect to risk of prostate cancer, and some large genes such as BRCA2 or MGMT had minimal coverage. However, the majority of selected tagSNPs in our study were also non-synonomous SNPs with potential functional prediction based on the in silico Polyphen program. Another limitation is the small number of African American men in this study that limited the statistical power to examine associations in this group.

Two potential issues that should be considered when pooling datasets of prostate cancer cases with different ascertainment periods are changes in the prevalence of PSA screening as well as a shift of Gleason score reading system over time (49–51). The increase of PSA screening over time would result theoretically in a higher proportion of prostate cancer cases diagnosed with minimal disease in the more recent study (49). To assess this issue, we compared the prevalence of self-reported PSA screening among cases and controls between the two studies. Although the frequency of PSA screening increased among controls from 34% to 58% when comparing study I (1993–1996) and study II (2002–2005), the frequency of PSA screening was similar between prostate cancer cases with different ascertainment periods, 73% and 71%, respectively. In addition the distribution of Gleason score 8 to 10 tumors was similar between the two studies, 9% and 10% respectively, although the prevalence of cases diagnosed with regional or distant stage tumors slightly declined from 26% to 18% when comparing study I (1993–1996) and study II (2002–2005). Another issue is the shift of Gleason score reading over time, with the tendency of pathologists to provide a higher Gleason score for prostate biopsy tumors in the more recent PSA screening era (50, 51). Although we did not assess this issue directly, as mentioned before the distribution of Gleason score was similar between the two studies, and the majority of our cases also received radical prostatectomy which usually corrects (either upgrades or downgrades) Gleason scores readings provided from the biopsies. Finally we stratified our data by study ascertainment period and run separate analysis for study I and II, respectively. We didn’t find any significant differences in ORs estimates obtained for each study separately, providing further reassurance that data from both studies are comparable.

In conclusion, we found no associations between DNA repair SNPs and overall risk of prostate cancer. Although, two SNPs in BRCA2 (rs144848, rs1801406) and two SNPs in ERCC2 (rs1799793, rs13181) showed some associations with Gleason score or tumor stage when comparing homozygous men carrying the minor vs. major allele, results were not statistically significantly different between clinically aggressive and non-aggressive tumors. Although our findings suggest that selected SNPs in DNA repair genes do not contribute to prostate cancer susceptibility, given the complexity of this pathway and its crucial role in maintenance of genomic stability a more comprehensive evaluation of tag SNPs, haplotypes, copy number variations and pathway-based analyses of all 150 genes in several DNA repair pathways, as well as exploration of gene-environment interactions may be warranted.

Acknowledgments

We are grateful to all men who participated in these studies for their time, effort and cooperation, and interviewers for their help with data collection. This work was supported by NIH grants R01-CA56678, R01-CA092579 and contract N01-CN-05230 from the National Cancer Institute. Additional support was provided by the Fred Hutchinson Cancer Research Center and the Intramural Program of the National Human Genome Research Institute. Ilir Agalliu was supported by funds from the Albert Einstein College of Medicine of Yeshiva University.

References

  • 1.American Cancer Society. Cancer Facts and Figures. Atlanta: 2009. [Google Scholar]
  • 2.Lichtenstein P, Holm NV, Verkasalo PK, Iliadou A, Kaprio J, Koskenvuo M, 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(2):78–85. doi: 10.1056/NEJM200007133430201. [DOI] [PubMed] [Google Scholar]
  • 3.Bostwick DG, Alexander EE, Singh R, Shan A, Qian J, Santella RM, et al. Antioxidant enzyme expression and reactive oxygen species damage in prostatic intraepithelial neoplasia and cancer. Cancer. 2000 Jul 1;89(1):123–34. [PubMed] [Google Scholar]
  • 4.Pathak SK, Sharma RA, Steward WP, Mellon JK, Griffiths TR, Gescher AJ. Oxidative stress and cyclooxygenase activity in prostate carcinogenesis: targets for chemopreventive strategies. Eur J Cancer. 2005;41(1):61–70. doi: 10.1016/j.ejca.2004.09.028. [DOI] [PubMed] [Google Scholar]
  • 5.Khandrika L, Kumar B, Koul S, Maroni P, Koul HK. Oxidative stress in prostate cancer. Cancer Lett. 2009 Sep 18;282(2):125–36. doi: 10.1016/j.canlet.2008.12.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Cooke MS, Evans MD, Dizdaroglu M, Lunec J. Oxidative DNA damage: mechanisms, mutation, and disease. Faseb J. 2003 Jul;17(10):1195–214. doi: 10.1096/fj.02-0752rev. [DOI] [PubMed] [Google Scholar]
  • 7.Sikka SC. Role of oxidative stress response elements and antioxidants in prostate cancer pathobiology and chemoprevention--a mechanistic approach. Curr Med Chem. 2003 Dec;10(24):2679–92. doi: 10.2174/0929867033456341. [DOI] [PubMed] [Google Scholar]
  • 8.Rybicki BA, Rundle A, Savera AT, Sankey SS, Tang D. Polycyclic aromatic hydrocarbon-DNA adducts in prostate cancer. Cancer Res. 2004 Dec 15;64(24):8854–9. doi: 10.1158/0008-5472.CAN-04-2323. [DOI] [PubMed] [Google Scholar]
  • 9.Rybicki BA, Nock NL, Savera AT, Tang D, Rundle A. Polycyclic aromatic hydrocarbon-DNA adduct formation in prostate carcinogenesis. Cancer Lett. 2006 Aug 8;239(2):157–67. doi: 10.1016/j.canlet.2005.07.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Tang D, Liu JJ, Rundle A, Neslund-Dudas C, Savera AT, Bock CH, et al. Grilled meat consumption and PhIP-DNA adducts in prostate carcinogenesis. Cancer Epidemiol Biomarkers Prev. 2007 Apr;16(4):803–8. doi: 10.1158/1055-9965.EPI-06-0973. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Christmann M, Tomicic MT, Roos WP, Kaina B. Mechanisms of human DNA repair: an update. Toxicology. 2003 Nov 15;193(1–2):3–34. doi: 10.1016/s0300-483x(03)00287-7. [DOI] [PubMed] [Google Scholar]
  • 12.Wood RD, Mitchell M, Sgouros J, Lindahl T. Human DNA repair genes. Science. 2001 Feb 16;291(5507):1284–9. doi: 10.1126/science.1056154. [DOI] [PubMed] [Google Scholar]
  • 13.Pierce AJ, Stark JM, Araujo FD, Moynahan ME, Berwick M, Jasin M. Double-strand breaks and tumorigenesis. Trends Cell Biol. 2001;11:S52–9. doi: 10.1016/s0962-8924(01)02149-3. [DOI] [PubMed] [Google Scholar]
  • 14.Jackson SP. Sensing and repairing DNA double-strand breaks. Carcinogenesis. 2002;23:687–96. doi: 10.1093/carcin/23.5.687. [DOI] [PubMed] [Google Scholar]
  • 15.Hu JJ, Mohrenweiser HW, Bell DA, Leadon SA, Miller MS. Symposium overview: genetic polymorphisms in DNA repair and cancer risk. Toxicol Appl Pharmacol. 2002 Nov 15;185(1):64–73. doi: 10.1006/taap.2002.9518. [DOI] [PubMed] [Google Scholar]
  • 16.Goode EL, Ulrich CM, Potter JD. Polymorphisms in DNA repair genes and associations with cancer risk. Cancer Epidemiol Biomarkers Prev. 2002 Dec;11(12):1513–30. [PubMed] [Google Scholar]
  • 17.Hu JJ, Hall MC, Grossman L, Hedayati M, McCullough DL, Lohman K, et al. Deficient nucleotide excision repair capacity enhances human prostate cancer risk. Cancer Res. 2004 Feb 1;64(3):1197–201. doi: 10.1158/0008-5472.can-03-2670. [DOI] [PubMed] [Google Scholar]
  • 18.Hung RJ, Hall J, Brennan P, Boffetta P. Genetic polymorphisms in the base excision repair pathway and cancer risk: a HuGE review. Am J Epidemiol. 2005 Nov 15;162(10):925–42. doi: 10.1093/aje/kwi318. [DOI] [PubMed] [Google Scholar]
  • 19.Xu J, Zheng SL, Turner A, Isaacs SD, Wiley KE, Hawkins GA, et al. Associations between hOGG1 sequence variants and prostate cancer susceptibility. Cancer Res. 2002 Apr 15;62(8):2253–7. [PubMed] [Google Scholar]
  • 20.van Gils CH, Bostick RM, Stern MC, Taylor JA. Differences in base excision repair capacity may modulate the effect of dietary antioxidant intake on prostate cancer risk: an example of polymorphisms in the XRCC1 gene. Cancer Epidemiol Biomarkers Prev. 2002 Nov;11(11):1279–84. [PubMed] [Google Scholar]
  • 21.Rybicki BA, Conti DV, Moreira A, Cicek M, Casey G, Witte JS. DNA repair gene XRCC1 and XPD polymorphisms and risk of prostate cancer. Cancer Epidemiol Biomarkers Prev. 2004 Jan;13(1):23–9. doi: 10.1158/1055-9965.epi-03-0053. [DOI] [PubMed] [Google Scholar]
  • 22.Ritchey JD, Huang WY, Chokkalingam AP, Gao YT, Deng J, Levine P, et al. Genetic variants of DNA repair genes and prostate cancer: a population-based study. Cancer Epidemiol Biomarkers Prev. 2005 Jul;14(7):1703–9. doi: 10.1158/1055-9965.EPI-04-0809. [DOI] [PubMed] [Google Scholar]
  • 23.Nock NL, Cicek MS, Li L, Liu X, Rybicki BA, Moreira A, et al. Polymorphisms in estrogen bioactivation, detoxification and oxidative DNA base excision repair genes and prostate cancer risk. Carcinogenesis. 2006 Sep;27(9):1842–8. doi: 10.1093/carcin/bgl022. [DOI] [PubMed] [Google Scholar]
  • 24.Chen L, Ambrosone CB, Lee J, Sellers TA, Pow-Sang J, Park JY. Association between polymorphisms in the DNA repair genes XRCC1 and APE1, and the risk of prostate cancer in white and black Americans. J Urol. 2006 Jan;175(1):108–12. doi: 10.1016/S0022-5347(05)00042-X. discussion 12. [DOI] [PubMed] [Google Scholar]
  • 25.Hirata H, Hinoda Y, Tanaka Y, Okayama N, Suehiro Y, Kawamoto K, et al. Polymorphisms of DNA repair genes are risk factors for prostate cancer. Eur J Cancer. 2007 Jan;43(2):231–7. doi: 10.1016/j.ejca.2006.11.005. [DOI] [PubMed] [Google Scholar]
  • 26.Hooker S, Bonilla C, Akereyeni F, Ahaghotu C, Kittles RA. NAT2 and NER genetic variants and sporadic prostate cancer susceptibility in African Americans. Prostate Cancer Prostatic Dis. 2008;11(4):349–56. doi: 10.1038/sj.pcan.4501027. [DOI] [PubMed] [Google Scholar]
  • 27.Stanford JL, Wicklund KG, McKnight B, Daling JR, Brawer MK. Vasectomy and risk of prostate cancer. Cancer Epidemiol Biomarkers Prev. 1999;8(10):881–6. [PubMed] [Google Scholar]
  • 28.Agalliu I, Salinas CA, Hansten PD, Ostrander EA, Stanford JL. Statin use and risk of prostate cancer: results from a population-based epidemiologic study. Am J Epidemiol. 2008 Aug 1;168(3):250–60. doi: 10.1093/aje/kwn141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Sunyaev S, Ramensky V, Koch I, Lathe W, 3rd, Kondrashov AS, Bork P. Prediction of deleterious human alleles. Hum Mol Genet. 2001 Mar 15;10(6):591–7. doi: 10.1093/hmg/10.6.591. [DOI] [PubMed] [Google Scholar]
  • 30.Sambrook J, Fritsch EF, Maniatis T. Isolation of high-molecular weight DNA from mammalian cells. In: Nolan C, editor. Molecular Cloning: A Laboratory Manual. Plainview, NY: Cold Spring Harbor Press; 1989. pp. 9.16–9.9. [Google Scholar]
  • 31.Ott J. Analysis of human genetic linkage. Baltimore: John Hopkins University Press; 1999. [Google Scholar]
  • 32.Breslow NE, Day NE. Statistical Methods in Cancer Research, Volume 1-The Analysis of Case-Control Studies. Lyon: International Agency for Research on Cancer; 1980. [PubMed] [Google Scholar]
  • 33.Klienbaum DG, Nizam A, Kupper L, Muller KE. Applied regression analysis and multivariate methods. 4. Pacific Grove, CA: Duxbury Press; 2007. [Google Scholar]
  • 34.Dubin N, Pasternack BS. Risk assessment for case-control subgroups by polychotomous logistic regression. Am J Epidemiol. 1986;123(6):1101–17. doi: 10.1093/oxfordjournals.aje.a114338. [DOI] [PubMed] [Google Scholar]
  • 35.The Breast Cancer Linkage C. Cancer risks in BRCA2 mutation carriers. J Natl Cancer Inst. 1999;91(15):1310–6. doi: 10.1093/jnci/91.15.1310. [DOI] [PubMed] [Google Scholar]
  • 36.Johannsson O, Loman N, Moller T, Kristoffersson U, Borg A, Olsson H. Incidence of malignant tumours in relatives of BRCA1 and BRCA2 germline mutation carriers. Eur J Cancer. 1999;35(8):1248–57. doi: 10.1016/s0959-8049(99)00135-5. [DOI] [PubMed] [Google Scholar]
  • 37.Eerola H, Pukkala E, Pyrhonen S, Blomqvist C, Sankila R, Nevanlinna H. Risk of cancer in BRCA1 and BRCA2 mutation-positive and -negative breast cancer families (Finland) Cancer Causes Control. 2001;12:739–46. doi: 10.1023/a:1011272919982. [DOI] [PubMed] [Google Scholar]
  • 38.Tulinius H, Olafsdottir GH, Sigvaldason H, Arason A, Barkardottir RB, Egilsson V, et al. The effect of a single BRCA2 mutation on cancer in Iceland. J Med Genet. 2002;39(7):457–62. doi: 10.1136/jmg.39.7.457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Bermejo JL, Hemminki K. Risk of cancer at sites other than the breast in Swedish families eligible for BRCA1 or BRCA2 mutation testing. Ann Oncol. 2004;15:1834–41. doi: 10.1093/annonc/mdh474. [DOI] [PubMed] [Google Scholar]
  • 40.van Asperen CJ, Brohet RM, Meijers-Heijboer EJ, Hoogerbrugge N, Verhoef S. Cancer risks in BRCA2 families: estimates for sites other than breast and ovary. J Med Genet. 2005;42:711–9. doi: 10.1136/jmg.2004.028829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Sigurdsson S, Thorlacius S, Tomasson J, Tryggvadottir L, Benediktsdottir K, Eyfjörd JE, et al. BRCA2 mutation in Icelandic prostate cancer patients. J Mol Med. 1997;75:758–61. doi: 10.1007/s001090050162. [DOI] [PubMed] [Google Scholar]
  • 42.Struewing JP, Hartge P, Wacholder S, Baker SM, Berlin M, McAdams M, et al. The risk of cancer associated with specific mutations of BRCA1 and BRCA2 among Ashkenazi Jews. N Engl J Med. 1997;336(20):1401–8. doi: 10.1056/NEJM199705153362001. [DOI] [PubMed] [Google Scholar]
  • 43.Kirchhoff T, Kauff ND, Mitra N, Nafa K, Huang H, Palmer C, et al. BRCA mutations and risk of prostate cancer in Ashkenazi Jews. Clin Cancer Res. 2004;10:2918–21. doi: 10.1158/1078-0432.ccr-03-0604. [DOI] [PubMed] [Google Scholar]
  • 44.Agalliu I, Gern R, Leanza S, Burk RD. Associations of high-grade prostate cancer with BRCA1 and BRCA2 founder mutations. Clin Cancer Res. 2009 Feb 1;15(3):1112–20. doi: 10.1158/1078-0432.CCR-08-1822. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Edwards SM, Kote-Jarai Z, Meitz J, Hamoudi R, Hope Q, Osin P, et al. Two percent of men with early-onset prostate cancer harbor germline mutations in the BRCA2 gene. Am J Hum Genet. 2003;72:1–12. doi: 10.1086/345310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Agalliu I, Karlins E, Kwon EM, Iwasaki LM, Diamond A, Ostrander EA, et al. Rare germline mutations in the BRCA2 gene are associated with early-onset prostate cancer. Br J Cancer. 2007 Sep 17;97(6):826–31. doi: 10.1038/sj.bjc.6603929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Ostrander EA, Udler MS. The role of the BRCA2 gene in susceptibility to prostate cancer revisited. Cancer Epidemiol Biomarkers Prev. 2008 Aug;17(8):1843–8. doi: 10.1158/1055-9965.EPI-08-0556. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Braithwaite E, Wu X, Wang Z. Repair of DNA lesions induced by polycyclic aromatic hydrocarbons in human cell-free extracts: involvement of two excision repair mechanisms in vitro. Carcinogenesis. 1998 Jul;19(7):1239–46. doi: 10.1093/carcin/19.7.1239. [DOI] [PubMed] [Google Scholar]
  • 49.Platz EA, De Marzo AM, Giovannucci E. Prostate cancer association studies: pitfalls and solutions to cancer misclassification in the PSA era. J Cell Biochem. 2004;91(3):553–71. doi: 10.1002/jcb.10700. [DOI] [PubMed] [Google Scholar]
  • 50.Kondylis FI, Moriarty RP, Bostwick D, Schellhammer PF. Prostate cancer grade assignment: the effect of chronological, interpretive and translation bias. J Urol. 2003 Oct;170(4 Pt 1):1189–93. doi: 10.1097/01.ju.0000085675.96097.76. [DOI] [PubMed] [Google Scholar]
  • 51.Albertsen PC, Hanley JA, Barrows GH, Penson DF, Kowalczyk PDH. Prostate cancer and the Will Rogers phenomenon. J Natl Cancer Inst. 2005;97(17):1248–52. doi: 10.1093/jnci/dji248. [DOI] [PubMed] [Google Scholar]

RESOURCES