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

Ilir Agalliu1, Erika M Kwon2, Claudia A. Salinas3,4,*, Joseph S. Koopmeiners3,5, Elaine A.Ostrander2, and Janet L. Stanford3,41 Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx,NY2 Cancer Genetics Branch, National Human Genome Research Institute, National Institutes ofHealth, Bethesda, MD3 Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA4 Department of Epidemiology, School of Public Health, University of Washington, Seattle, WA5 Department of Biostatistics, School of Public Health, University of Washington, Seattle, WA

AbstractObjective—DNA repair pathways are crucial to prevent accumulation of DNA damage andmaintain genomic stability. Alterations of this pathway have been reported in many cancers. Anincrease in oxidative DNA damage or decrease of DNA repair capacity with aging or due to germlinegenetic variation may affect prostate cancer risk.

Methods—Pooled data from two population-based studies (1,457 cases and 1,351 controls) wereanalyzed 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 exploredwhether 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. Risksdid not vary either by smoking or by family history of prostate cancer. Although, two SNPs inBRCA2 (rs144848, rs1801406) and two SNPs in ERCC2 (rs1799793, rs13181) showed strongerassociations with high Gleason score or advanced stage tumors when comparing homozygous mencarrying the minor vs. major allele, results were not statistically significantly different betweenclinically aggressive and non-aggressive tumors.

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

KeywordsDNA repair; SNPs; prostate cancer; case-control study; genetic variation

Address correspondence to: Janet L. Stanford, Ph.D., Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, 1100Fairview Ave N., M4-B847, Seattle, WA 98109-1024, Phone: (206) 667-2715, Fax: (206) 667-2717, jstanfor@fhcrc.org.*Current affiliation of Claudia Salinas is: Division of Hematology and Oncology, University of Michigan Cancer Center, Ann Arbor,MI 48109.

NIH Public AccessAuthor ManuscriptCancer Causes Control. Author manuscript; available in PMC 2011 February 1.

Published in final edited form as:Cancer Causes Control. 2010 February ; 21(2): 289–300. doi:10.1007/s10552-009-9461-5.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

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

DNA repair mechanisms are important pathways in removal of oxidative DNA compounds orDNA 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 removessingle base modifications including single strand breaks, oxidative DNA damage and non-bulky adducts, where as nucleotide excision repair removes larger lesions, which often resultfrom environmental exposures such as smoking, UV radiation or external carcinogens (11,12). Alkyltransferases directly reverse DNA damage by transferring alkyl groups fromdamaged DNA onto the transferase enzyme (11). Finally, double-strand DNA breaks arerepaired through complex mechanisms including homologous recombination and end-joiningrepair 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 DNAadducts or lesions, may result in enhanced cancer risk (15–18).

Several epidemiologic studies have examined associations between single-nucleotidepolymorphisms (SNPs) in DNA repair genes, mostly non-synonymous SNPs with potentialfunctional significance, and risk of prostate cancer (17,19–26). However, results have beeninconsistent 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 aswell as limited power to detect modest associations. The objective of our study was to examineassociations between genetic variants in DNA repair genes and risk of prostate cancer in a largepopulation-based case-control dataset, with detailed information on demographic and lifestylefactors, screening as well as clinical features of the disease. In addition, we were interested inexploring whether SNP-prostate cancer associations varied according to smoking status, familyhistory of prostate cancer, or clinical features of this cancer.

Materials and MethodsStudy 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 prostatecancer with ascertainment periods from 1993 to 1996 and 2002 to 2005 that have beenpreviously described (27,28). Incident cases with histologically confirmed adeno-carcinomaof the prostate were 35 to 74 years old at diagnosis, and were identified via the Seattle-PugetSound Surveillance, Epidemiology, and End Results (SEER) cancer registry. The SEER cancerregistry also provided information on clinical characteristics of prostate cancer includingGleason score, tumor stage, and serum prostate-specific antigen (PSA) levels at diagnosis, aswell as primary treatment for prostate cancer. Controls were men without a self-reportedphysician’s diagnosis of prostate cancer, identified via random digit dialing, frequency matched

Agalliu et al. Page 2

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

to cases by 5-year age groups, and recruited evenly throughout the ascertainment period of thecases. Both studies were approved by the Institutional Review Board (IRB) of the FredHutchinson Cancer Research Center, and written informed consent was obtained from all studyparticipants. Genotyping was approved by the IRB of the National Human Genome ResearchInstitute.

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

Selection of SNPs and GenotypingSNPs in DNA repair genes were selected for genotyping in this study based on the followingcriteria: (a) functional significance, as assessed by potential effects on enzyme level activityor projected influence on DNA adducts levels, (b) previously reported association(s) withprostate cancer risk in earlier studies, (c) tagSNPs with a minor allele frequency ≥5% selectedfrom 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 SNPsin 9 DNA repair genes genotyped in our study were non-synonomous SNP (n=13) followedby 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 ofvariation across large genes such as BRCA2 or MGMT is small. SNPs with prior evidence fromassociation analyses reported from other studies were included, except for the SNPs in themismatch DNA repair pathway, which were part of another analysis. Potential functionalprediction of SNPs that resulted in amino-acid changes was assessed using in silicoclassification program Polyphen (29).

Genomic DNA was purified from peripheral lymphocytes using standard protocols (30). TheApplied Biosystems (ABI) SNPlex™ Genotyping system was used for genotyping andGeneMapper software was used for calling alleles. Discrimination of the specific SNP allelewas carried out with the Applied Biosystems 3730xl DNA analyzer and was based on thepresence of a unique sequence assigned to the original allele-specific oligonucleotide. Qualitycontrol included genotyping of 140 blind duplicate samples that were distributed across allgenotyping batches. Each batch of DNA aliquots that were genotyped incorporated similarnumbers of cases and controls samples collected in a specific calendar year (e.g. 2003) andsamples were distributed equally across the entire genotyping plates. The laboratory personnelthat 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, andtwo others (ERCC4 rs1799802, MGMT rs2308318) were monomorphic in our studypopulation. One SNP (rs1052133) in the OGG1 had a drop-out rate of 9.5%, and thus wasexcluded from all statistical analyses. The remaining 28 SNPs in 9 DNA repair genes with anaverage drop-out of 0.6% and an average genotype concordance of 99.7% across 140 blindduplicate samples were included in statistical analyses.

Agalliu et al. Page 3

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Statistical AnalysesDeviation of genotype frequencies from Hardy-Weinberg Equilibrium (HWE) amongCaucasian and African-American controls considered separately was assessed by χ2-tests(31). Unconditional logistic regression was used to examine associations between SNPgenotypes and prostate cancer risk among Caucasian and African-American men and tocompute odds ratios (OR) and 95% confidence intervals (CI) (32). Confounding by establishedand potential risk factors for prostate cancer was assessed for each genotype separately, fittingmodels using each main effect and then evaluating the change in parameter estimates of theSNP genotypes when other variables entered the models one at a time. All analyses wereadjusted for age at diagnosis for cases and at reference date for controls. Further adjustmentfor 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 includedin the final models. Both additive and dominant genetic models were considered for eachvariant allele depending on the distribution of genotypes. Likelihood ratio-based test statisticswere used to identify statistically significant associations between SNP genotypes and prostatecancer 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 dominantmodels were fit for all SNPs and the minimum of these p-values were kept for each SNP. Thep-values were then ordered to approximate the null distribution of the order statistics for thep-values, starting from the smallest to the largest. The original p-values were also ordered andpermutation p-values were calculated by comparing the ordered p-values to the null distributionfor the appropriate order statistic. Permutation p-values can be interpreted as the probabilityof observing a p-value less than or equal to what was observed for the given order statisticunder the null hypothesis of no association with disease for any of the 28 SNPs. The permutationapproach to approximating the null distribution of the order statistics will be valid regardlessof any correlation between the SNPs. A SNP was considered to be significantly associated withprostate 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 prostatecancer (yes vs. no) or between base- or nucleotide excision repair genes and either smokingstatus 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 andcomparing the fully saturated model containing the main effects and interaction term with thereduced model containing only the main effects using a likelihood ratio test to determinestatistical significance of the interaction effects (33). Lastly, we also examined associationsbetween SNP genotypes and clinical characteristics of prostate cancer. With respect to theGleason 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 diagnosedwith regional or distant stage were compared to men with localized stage at diagnosis. Thefrequency of genotypes for DNA repair SNPs in each group of cases was compared to thefrequency of genotypes in controls using polytomous logistic regression models (34); thesefinal models were adjusted for age and prostate cancer screening history.

ResultsSelected characteristics of prostate cancer cases and controls, stratified by race, are presentedin 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 thiscould have been due to small number of cases and controls in this category. In both racialgroups, cases were more likely than controls to report a first-degree family history of prostate

Agalliu et al. Page 4

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

cancer and to have undergone PSA screening within the 5 year-period before diagnosis orreference date. Cases and controls were similar with respect to other factors including bodymass index, education and smoking. With respect to clinical features, the majority of prostatecancer cases were diagnosed with localized stage or Gleason score 2–6 or 7(3+4) cancers andthe distribution of Gleason score and tumor stage was similar between cases ascertained in twodifferent periods: 1993–1996 and 2002–2005 (data not shown).

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

Next we explored interactions between DNA repair SNPs and first-degree family history ofprostate cancer (yes vs. no) and smoking (both smoking status and lifetime pack-years ofsmoking) 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 clinicalcharacteristics 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) inBRCA2 and two SNP (rs1799793, rs13181) in ERCC2 showed some associations with Gleasonscore or tumor stage in single SNP analyses. For BRCA2 rs144848, although men with the GGgenotype had an OR of 1.83 (95% CI 1.09–3.08) of high Gleason score tumors in comparisonto men with the TT genotype, the risk estimate was not statistically significantly different incomparison 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 highGleason score tumors in comparison to men homozygous for the major allele (ORs of 0.48 and0.54, respectively) the risk estimates between higher and lower Gleason score tumors were notstatistically significantly different. Similar findings of no significant differences were alsoobserved in relation to tumor stage when data were stratified according to localized tumorsversus regional/distant tumors. No associations were observed between the remaining 24 SNPsin DNA repair genes and any of the clinical characteristics of prostate cancer.

DiscussionIn this population-based study we examined associations between 28 SNPs in 9 DNA repairgenes and prostate cancer risk among 1,457 cases and 1,351 controls pooled from two priorstudies conducted in King County, Washington. No associations were observed between anyof these SNPs and overall risk of prostate cancer, after adjusting for age and multiplecomparisons. Risks were not different according to family history of prostate cancer or bysmoking (either smoking status or pack-years of smoking) with mostly null associations. Withrespect to clinical characteristics of prostate cancer, two SNPs in BRCA2 (rs144848,rs1801406) and two SNP in ERCC2 (rs1799793, rs13181) showed some associations withGleason score and tumor stage in single SNP analysis; however ORs were not statisticallysignificantly different between lower and higher Gleason score tumors or between localizedversus regional or distant stage tumors.

Agalliu et al. Page 5

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

In relation to BRCA2 SNPs, although both rs144848 (Asn372His) and rs1801406 (Lys1132Lys) 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 Gleason2–6 or 7(3+4) tumors. However, none of the above SNPs in BRCA2 were associated with tumorstage 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 ofpopulations 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 forfamily history (45,46) all have reported that men who carry protein-truncating BRCA2mutations have increased risk of prostate cancer with relative risk (RRs) ranging from 2.0 to23.0 depending on several factors including study population, case ascertainment, age atprostate cancer diagnosis and familial history of several cancer such as prostate, breast andovary. Unlike SNPs, which have an unclear role in protein function, the above deleteriousmutations in BRCA2 result in an earlier truncation of the BRCA2 protein and thus are morelikely to be linked to prostate cancer susceptibility; however given their very low prevalencein the general population (~0.1%) it is estimated that <1% of sporadic prostate cancers can beattributed to these disease-associated BRCA2 mutations (47).

With respect to ERCC2 (XPD) SNPs, we found no association with overall risk of prostatecancer; however there was an inverse association between two SNPs: rs1799793 (Asp312Asn)and rs13181 (Lys751Glu) and clinical features of prostate cancer including Gleason score andtumor stage. For both of these two non-synonymous SNPs although men homozygous for theminor allele had a stronger reduction in risk of higher Gleason score or regional or distanttumor stage in comparison to men with the more common genotype, these ORs were notstatistically significant for those obtained for low Gleason score (2–6 or 7=3+4) and localizedstage 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-excisionrepair (NER) pathway (12). This pathway is responsible for removal of DNA bulky lesionsthat 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 haveexamined associations between NER pathway and prostate cancer risk (17,21). Hu andcolleagues (17) examined associations between NER capacity in isolated prostate tissue asmeasured in a host-cell reaction assay, and prostate cancer risk is a small clinic-based study of140 prostate cancer patients and 96 controls. They reported a 2.6-fold increased risk (95%CI1.2–6.0) of prostate cancer for men in the lowest quartile of NER capacity in comparison tothose in the highest quartile (17), however, they did not consider genetic variants in NERpathway genes including ERCC2. In another study, Rybicki and colleagues (21) examinedassociations between the same two variants in ERCC2 (rs1799793, rs13181) that we evaluatedand risk of prostate cancer in a family-based study of 637 cases and 480 brother controls (thestudy population was primarily Caucasian) and reported a positive association betweenrs1799793 (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). Noassociation was observed for ERCC2 rs13181 in that study (21), nor risks differed by clinicalcharacteristics 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 resultscould 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 overallrisk of prostate cancer or clinical characteristics of this disease in our study population, althoughprior studies have reported positive associations between risk of prostate cancer and geneticvariants 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

Agalliu et al. Page 6

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

(95% CI 1.19–8.73) for sporadic prostate cancer risk associated with the GG genotype vs. CCgenotype for OGG1 rs1052133 (Ser326Cys). However that SNP was excluded from ouranalysis due to high frequency of drop-outs. In relation to XRCC1, two studies of Caucasianpopulations (20,21) reported no associations between prostate cancer and rs25487(Arg399Gln), rs1799782 (Arg194Trp) or rs25489 (Arg280His), which is similar to ourfindings. 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, reportedpositive associations between risk of prostate cancer and MGMT Leu84Phe (22), XPCLys939Gln (25) as well as XRCC1 Arg194Trp (22)(25). However the allele frequencies ofthese 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 wheninterpreting the results. Strengths of the current study include the population-based design, thesample size, and the availability of information on potential confounding variables, as well asthe availability of clinical data on prostate cancer cases. A limitation of our study is that weexamined only a small number of SNPs in DNA repair genes with respect to risk of prostatecancer, and some large genes such as BRCA2 or MGMT had minimal coverage. However, themajority of selected tagSNPs in our study were also non-synonomous SNPs with potentialfunctional prediction based on the in silico Polyphen program. Another limitation is the smallnumber of African American men in this study that limited the statistical power to examineassociations in this group.

Two potential issues that should be considered when pooling datasets of prostate cancer caseswith different ascertainment periods are changes in the prevalence of PSA screening as wellas a shift of Gleason score reading system over time (49–51). The increase of PSA screeningover time would result theoretically in a higher proportion of prostate cancer cases diagnosedwith minimal disease in the more recent study (49). To assess this issue, we compared theprevalence 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% whencomparing study I (1993–1996) and study II (2002–2005), the frequency of PSA screeningwas 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 betweenthe two studies, 9% and 10% respectively, although the prevalence of cases diagnosed withregional 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 readingover time, with the tendency of pathologists to provide a higher Gleason score for prostatebiopsy tumors in the more recent PSA screening era (50,51). Although we did not assess thisissue directly, as mentioned before the distribution of Gleason score was similar between thetwo studies, and the majority of our cases also received radical prostatectomy which usuallycorrects (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 studyI and II, respectively. We didn’t find any significant differences in ORs estimates obtained foreach 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 prostatecancer. Although, two SNPs in BRCA2 (rs144848, rs1801406) and two SNPs in ERCC2(rs1799793, rs13181) showed some associations with Gleason score or tumor stage whencomparing homozygous men carrying the minor vs. major allele, results were not statisticallysignificantly different between clinically aggressive and non-aggressive tumors. Although ourfindings suggest that selected SNPs in DNA repair genes do not contribute to prostate cancersusceptibility, given the complexity of this pathway and its crucial role in maintenance ofgenomic stability a more comprehensive evaluation of tag SNPs, haplotypes, copy number

Agalliu et al. Page 7

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

variations and pathway-based analyses of all 150 genes in several DNA repair pathways, aswell as exploration of gene-environment interactions may be warranted.

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

References1. American Cancer Society. Cancer Facts and Figures. Atlanta: 2009.2. 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, andFinland. N Engl J Med 2000;343(2):78–85. [PubMed: 10891514]

3. Bostwick DG, Alexander EE, Singh R, Shan A, Qian J, Santella RM, et al. Antioxidant enzymeexpression and reactive oxygen species damage in prostatic intraepithelial neoplasia and cancer.Cancer 2000 Jul 1;89(1):123–34. [PubMed: 10897009]

4. Pathak SK, Sharma RA, Steward WP, Mellon JK, Griffiths TR, Gescher AJ. Oxidative stress andcyclooxygenase activity in prostate carcinogenesis: targets for chemopreventive strategies. Eur JCancer 2005;41(1):61–70. [PubMed: 15617991]

5. Khandrika L, Kumar B, Koul S, Maroni P, Koul HK. Oxidative stress in prostate cancer. Cancer Lett2009 Sep 18;282(2):125–36. [PubMed: 19185987]

6. Cooke MS, Evans MD, Dizdaroglu M, Lunec J. Oxidative DNA damage: mechanisms, mutation, anddisease. Faseb J 2003 Jul;17(10):1195–214. [PubMed: 12832285]

7. Sikka SC. Role of oxidative stress response elements and antioxidants in prostate cancer pathobiologyand chemoprevention--a mechanistic approach. Curr Med Chem 2003 Dec;10(24):2679–92. [PubMed:14529458]

8. Rybicki BA, Rundle A, Savera AT, Sankey SS, Tang D. Polycyclic aromatic hydrocarbon-DNAadducts in prostate cancer. Cancer Res 2004 Dec 15;64(24):8854–9. [PubMed: 15604244]

9. Rybicki BA, Nock NL, Savera AT, Tang D, Rundle A. Polycyclic aromatic hydrocarbon-DNA adductformation in prostate carcinogenesis. Cancer Lett 2006 Aug 8;239(2):157–67. [PubMed: 16154258]

10. Tang D, Liu JJ, Rundle A, Neslund-Dudas C, Savera AT, Bock CH, et al. Grilled meat consumptionand PhIP-DNA adducts in prostate carcinogenesis. Cancer Epidemiol Biomarkers Prev 2007 Apr;16(4):803–8. [PubMed: 17416774]

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. [PubMed: 14599765]

12. Wood RD, Mitchell M, Sgouros J, Lindahl T. Human DNA repair genes. Science 2001 Feb 16;291(5507):1284–9. [PubMed: 11181991]

13. Pierce AJ, Stark JM, Araujo FD, Moynahan ME, Berwick M, Jasin M. Double-strand breaks andtumorigenesis. Trends Cell Biol 2001;11:S52–9. [PubMed: 11684443]

14. Jackson SP. Sensing and repairing DNA double-strand breaks. Carcinogenesis 2002;23:687–96.[PubMed: 12016139]

15. Hu JJ, Mohrenweiser HW, Bell DA, Leadon SA, Miller MS. Symposium overview: geneticpolymorphisms in DNA repair and cancer risk. Toxicol Appl Pharmacol 2002 Nov 15;185(1):64–73. [PubMed: 12460738]

16. Goode EL, Ulrich CM, Potter JD. Polymorphisms in DNA repair genes and associations with cancerrisk. Cancer Epidemiol Biomarkers Prev 2002 Dec;11(12):1513–30. [PubMed: 12496039]

17. Hu JJ, Hall MC, Grossman L, Hedayati M, McCullough DL, Lohman K, et al. Deficient nucleotideexcision repair capacity enhances human prostate cancer risk. Cancer Res 2004 Feb 1;64(3):1197–201. [PubMed: 14871857]

Agalliu et al. Page 8

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

18. Hung RJ, Hall J, Brennan P, Boffetta P. Genetic polymorphisms in the base excision repair pathwayand cancer risk: a HuGE review. Am J Epidemiol 2005 Nov 15;162(10):925–42. [PubMed:16221808]

19. Xu J, Zheng SL, Turner A, Isaacs SD, Wiley KE, Hawkins GA, et al. Associations between hOGG1sequence variants and prostate cancer susceptibility. Cancer Res 2002 Apr 15;62(8):2253–7.[PubMed: 11956079]

20. van Gils CH, Bostick RM, Stern MC, Taylor JA. Differences in base excision repair capacity maymodulate the effect of dietary antioxidant intake on prostate cancer risk: an example ofpolymorphisms in the XRCC1 gene. Cancer Epidemiol Biomarkers Prev 2002 Nov;11(11):1279–84.[PubMed: 12433703]

21. Rybicki BA, Conti DV, Moreira A, Cicek M, Casey G, Witte JS. DNA repair gene XRCC1 and XPDpolymorphisms and risk of prostate cancer. Cancer Epidemiol Biomarkers Prev 2004 Jan;13(1):23–9. [PubMed: 14744728]

22. Ritchey JD, Huang WY, Chokkalingam AP, Gao YT, Deng J, Levine P, et al. Genetic variants ofDNA repair genes and prostate cancer: a population-based study. Cancer Epidemiol Biomarkers Prev2005 Jul;14(7):1703–9. [PubMed: 16030105]

23. Nock NL, Cicek MS, Li L, Liu X, Rybicki BA, Moreira A, et al. Polymorphisms in estrogenbioactivation, detoxification and oxidative DNA base excision repair genes and prostate cancer risk.Carcinogenesis 2006 Sep;27(9):1842–8. [PubMed: 16569655]

24. Chen L, Ambrosone CB, Lee J, Sellers TA, Pow-Sang J, Park JY. Association betweenpolymorphisms in the DNA repair genes XRCC1 and APE1, and the risk of prostate cancer in whiteand black Americans. J Urol 2006 Jan;175(1):108–12. discussion 12. [PubMed: 16406883]

25. Hirata H, Hinoda Y, Tanaka Y, Okayama N, Suehiro Y, Kawamoto K, et al. Polymorphisms of DNArepair genes are risk factors for prostate cancer. Eur J Cancer 2007 Jan;43(2):231–7. [PubMed:17196815]

26. Hooker S, Bonilla C, Akereyeni F, Ahaghotu C, Kittles RA. NAT2 and NER genetic variants andsporadic prostate cancer susceptibility in African Americans. Prostate Cancer Prostatic Dis 2008;11(4):349–56. [PubMed: 18026184]

27. Stanford JL, Wicklund KG, McKnight B, Daling JR, Brawer MK. Vasectomy and risk of prostatecancer. Cancer Epidemiol Biomarkers Prev 1999;8(10):881–6. [PubMed: 10548316]

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.[PubMed: 18556686]

29. Sunyaev S, Ramensky V, Koch I, Lathe W 3rd, Kondrashov AS, Bork P. Prediction of deleterioushuman alleles. Hum Mol Genet 2001 Mar 15;10(6):591–7. [PubMed: 11230178]

30. Sambrook, J.; Fritsch, EF.; Maniatis, T. Isolation of high-molecular weight DNA from mammaliancells. In: Nolan, C., editor. Molecular Cloning: A Laboratory Manual. Plainview, NY: Cold SpringHarbor Press; 1989. p. 9.16-9.9.

31. Ott, J. Analysis of human genetic linkage. Baltimore: John Hopkins University Press; 1999.32. Breslow, NE.; Day, NE. Statistical Methods in Cancer Research, Volume 1-The Analysis of Case-

Control Studies. Lyon: International Agency for Research on Cancer; 1980.33. Klienbaum, DG.; Nizam, A.; Kupper, L.; Muller, KE. Applied regression analysis and multivariate

methods. 4. Pacific Grove, CA: Duxbury Press; 2007.34. Dubin N, Pasternack BS. Risk assessment for case-control subgroups by polychotomous logistic

regression. Am J Epidemiol 1986;123(6):1101–17. [PubMed: 3706280]35. The Breast Cancer Linkage C. Cancer risks in BRCA2 mutation carriers. J Natl Cancer Inst 1999;91

(15):1310–6. [PubMed: 10433620]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. [PubMed: 10615237]

37. Eerola H, Pukkala E, Pyrhonen S, Blomqvist C, Sankila R, Nevanlinna H. Risk of cancer in BRCA1and BRCA2 mutation-positive and -negative breast cancer families (Finland). Cancer Causes Control2001;12:739–46. [PubMed: 11562114]

Agalliu et al. Page 9

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

38. Tulinius H, Olafsdottir GH, Sigvaldason H, Arason A, Barkardottir RB, Egilsson V, et al. The effectof a single BRCA2 mutation on cancer in Iceland. J Med Genet 2002;39(7):457–62. [PubMed:12114473]

39. Bermejo JL, Hemminki K. Risk of cancer at sites other than the breast in Swedish families eligiblefor BRCA1 or BRCA2 mutation testing. Ann Oncol 2004;15:1834–41. [PubMed: 15550590]

40. van Asperen CJ, Brohet RM, Meijers-Heijboer EJ, Hoogerbrugge N, Verhoef S. Cancer risks inBRCA2 families: estimates for sites other than breast and ovary. J Med Genet 2005;42:711–9.[PubMed: 16141007]

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. [PubMed:9383000]

42. Struewing JP, Hartge P, Wacholder S, Baker SM, Berlin M, McAdams M, et al. The risk of cancerassociated with specific mutations of BRCA1 and BRCA2 among Ashkenazi Jews. N Engl J Med1997;336(20):1401–8. [PubMed: 9145676]

43. Kirchhoff T, Kauff ND, Mitra N, Nafa K, Huang H, Palmer C, et al. BRCA mutations and risk ofprostate cancer in Ashkenazi Jews. Clin Cancer Res 2004;10:2918–21. [PubMed: 15131025]

44. Agalliu I, Gern R, Leanza S, Burk RD. Associations of high-grade prostate cancer with BRCA1 andBRCA2 founder mutations. Clin Cancer Res 2009 Feb 1;15(3):1112–20. [PubMed: 19188187]

45. Edwards SM, Kote-Jarai Z, Meitz J, Hamoudi R, Hope Q, Osin P, et al. Two percent of men withearly-onset prostate cancer harbor germline mutations in the BRCA2 gene. Am J Hum Genet2003;72:1–12. [PubMed: 12474142]

46. Agalliu I, Karlins E, Kwon EM, Iwasaki LM, Diamond A, Ostrander EA, et al. Rare germlinemutations in the BRCA2 gene are associated with early-onset prostate cancer. Br J Cancer 2007 Sep17;97(6):826–31. [PubMed: 17700570]

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. [PubMed: 18708369]

48. Braithwaite E, Wu X, Wang Z. Repair of DNA lesions induced by polycyclic aromatic hydrocarbonsin human cell-free extracts: involvement of two excision repair mechanisms in vitro. Carcinogenesis1998 Jul;19(7):1239–46. [PubMed: 9683183]

49. Platz EA, De Marzo AM, Giovannucci E. Prostate cancer association studies: pitfalls and solutionsto cancer misclassification in the PSA era. J Cell Biochem 2004;91(3):553–71. [PubMed: 14755685]

50. Kondylis FI, Moriarty RP, Bostwick D, Schellhammer PF. Prostate cancer grade assignment: theeffect of chronological, interpretive and translation bias. J Urol 2003 Oct;170(4 Pt 1):1189–93.[PubMed: 14501722]

51. Albertsen PC, Hanley JA, Barrows GH, Penson DF, Kowalczyk PDH. Prostate cancer and the WillRogers phenomenon. J Natl Cancer Inst 2005;97(17):1248–52. [PubMed: 16145045]

Agalliu et al. Page 10

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 11

Tabl

e 1

Sele

cted

cha

ract

eris

tics o

f pro

stat

e ca

ncer

cas

es a

nd c

ontro

ls g

enot

yped

for S

NPs

in D

NA

repa

ir ge

nes,

by ra

ce

Cau

casi

ans

Afr

ican

- A

mer

ican

s

Cha

ract

eris

ticC

ases

(N=1

,308

)C

ontr

ols (

N=1

,266

)P-

valu

e†C

ases

(N=1

49)

Con

trol

s (N

=85)

P-va

lue†

Age

(yea

rs);

n (%

)0.

360.

0002

 35

–49

102

(7.8

)10

7 (8

.5)

16 (1

0.7)

19 (2

2.4)

 50

–54

188

(14.

4)17

8 (1

4.1)

26 (1

7.5)

31 (3

6.5)

 55

–59

325

(24.

9)34

3 (2

7.1)

32 (2

1.5)

15 (1

7.7)

 60

–64

395

(30.

2)33

4 (2

6.3)

38 (2

5.5)

14 (1

6.5)

 65

–69

153

(11.

7)16

0 (1

2.6)

24 (1

6.1)

4 (4

.7)

 70

–74

145

(11.

1)14

4 (1

1.4)

13 (8

.7)

2 (2

.4)

Firs

t-deg

ree

fam

ily h

isto

ry o

fpr

osta

te c

ance

r; n

(%)

<0.0

001

0.04

 N

o1,

025

(78.

4)1,

125

(88.

9)11

9 (7

9.9)

75 (8

8.2)

 Y

es28

3 (2

1.6)

141

(11.

1)30

(20.

1)10

(11.

8)

Bod

y m

ass i

ndex

(BM

I), k

g/m

2 n(%

)0.

340.

39

 <

25.0

429

(32.

8)38

9 (3

0.7)

38 (2

5.5)

21 (2

4.7)

 25

.0–2

9.9

637

(48.

7)61

8 (4

8.8)

69 (4

6.1)

33 (3

8.8)

 ≥

3024

2 (1

8.5)

259

(20.

5)42

(28.

2)31

(36.

5)

Smok

ing

stat

us; n

(%)

0.13

0.64

 N

on sm

oker

522

(39.

9)54

1 (4

2.7)

63 (4

2.3)

31 (3

6.5)

 Fo

rmer

smok

er63

1 (4

8.2)

561

(44.

3)45

(30.

2)30

(35.

3)

 C

urre

nt sm

oker

155

(11.

9)16

4 (1

3.0)

41 (2

7.5)

24 (2

8.2)

Educ

atio

n; n

(%)

0.85

0.09

 H

igh

scho

ol o

r les

s23

0 (1

7.6)

237

(18.

7)57

(38.

3)24

(28.

2)

 So

me

colle

ge32

4 (2

4.8)

300

(23.

7)49

(32.

9)23

(27.

1)

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 12

Cau

casi

ans

Afr

ican

- A

mer

ican

s

Cha

ract

eris

ticC

ases

(N=1

,308

)C

ontr

ols (

N=1

,266

)P-

valu

e†C

ases

(N=1

49)

Con

trol

s (N

=85)

P-va

lue†

 C

olle

ge d

egre

e36

9 (2

8.2)

354

(28.

0)21

(14.

1)21

(24.

7)

 G

radu

ate

degr

ee38

5 (2

9.4)

375

(29.

6)22

(14.

7)17

(20.

0)

Pros

tate

can

cer s

cree

ning

hist

ory‡

; n (%

)<0

.000

10.

002

 N

one

135

(10.

3)16

8 (1

3.3)

22 (1

4.8)

15 (1

7.7)

 D

RE

only

223

(17.

1)48

2 (3

8.1)

34 (2

2.8)

36 (4

2.3)

 PS

A95

0 (7

2.6)

616

(48.

7)93

(62.

4)34

(40.

0)

Gle

ason

scor

e, n

(%)

 2–

466

(5.1

)5

(3.4

)

 5–

668

1 (5

2.1)

61 (4

0.9)

 7

(3+4

)35

5 (2

7.1)

53 (3

5.6)

 7

(4+3

)76

(5.8

)14

(9.4

)

 8–

1012

6 (9

.6)

14 (9

.4)

 M

issi

ng4

(0.3

)2

(1.3

)

Stag

e of

can

cer,

n (%

)

 Lo

caliz

ed1,

022

(78.

1)11

8 (7

9.2)

 R

egio

nal

254

(19.

4)26

(17.

5)

 D

ista

nt32

(2.5

)5

(3.4

)

PSA

val

ueξ

(ng/

ml);

n (%

)<0

.000

1<0

.000

1

 0.

01–3

.917

8 (1

3.6)

1,17

5 (9

2.8)

11 (7

.4)

78 (9

1.8)

 4.

0–9.

970

5 (5

3.9)

74 (5

.9)

91 (6

1.1)

6 (7

.1)

 10

.0–1

9.9

188

(14.

4)15

(1.2

)19

(12.

8)1

(1.2

)

 ≥

2011

8 (9

.0)

2 (0

.2)

20 (1

3.4)

-

 M

issi

ng11

9 (9

.1)

-8

(5.4

)-

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 13† C

hi-s

quar

e p-

valu

e;

‡ Pros

tate

can

cer s

cree

ning

his

tory

with

in th

e 5

year

s bef

ore

pros

tate

can

cer d

iagn

osis

or r

efer

ence

dat

e.

ξ Seru

m P

SA v

alue

at p

rost

ate

canc

er d

iagn

osis

for c

ases

and

at i

nter

view

for c

ontro

ls A

bbre

viat

ions

: DR

E - d

igita

l rec

tal e

xam

inat

ion;

PSA

- pr

osta

te sp

ecifi

c an

tigen

.

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 14

Tabl

e 2

Ass

ocia

tions

bet

wee

n SN

Ps in

DN

A re

pair

gene

s and

pro

stat

e ca

ncer

risk

, by

race

Cau

casi

ans

Afr

ican

-Am

eric

an

Cas

esC

ontr

ols

Cas

esC

ontr

ols

Gen

eSN

PG

enot

ype

N%

N%

OR

†95

% C

IPe

rmut

ed P

*N

%N

%O

R†

95%

CI

Perm

uted

P*

APEX

1rs

1320

150

AA

424

33.5

415

33.4

1.00

-0.

4257

40.1

3746

.81.

00-

0.84

AG

612

48.4

615

49.5

0.97

0.82

–1.1

665

45.8

3341

.81.

430.

77–2

.66

GG

229

18.1

212

17.1

1.06

0.84

–1.3

320

14.1

911

.41.

480.

58–3

.79

AG

+ G

G84

166

.582

766

.61.

000.

84–1

.18

8559

.942

53.2

1.44

0.80

–2.5

9

APEX

1rs

2275

007

GG

420

33.4

415

33.5

1.00

-0.

7961

43.0

3746

.81.

00-

0.53

AG

608

48.3

614

49.5

0.98

0.82

–1.1

762

43.7

3139

.21.

400.

75–2

.62

AA

230

18.3

211

17.0

1.08

0.85

–1.3

619

13.4

1113

.91.

080.

44–2

.65

AG

+ A

A83

866

.682

566

.51.

010.

85–1

.19

8157

.042

53.2

1.32

0.73

–2.3

6

BRC

A2rs

1448

48TT

655

51.6

654

52.6

1.00

-0.

5210

473

.259

74.7

1.00

-0.

36

GT

498

39.4

500

40.2

0.99

0.84

–1.1

736

25.4

1822

.80.

980.

49–1

.95

GG

116

9.1

897.

21.

300.

98–1

.75

21.

42

2.5

0.48

0.06

–3.8

5

GT

+ G

G61

448

.558

947

.41.

040.

89–1

.22

3826

.820

25.3

0.93

0.48

–1.8

1

BRC

A2rs

1586

9A

A77

060

.878

763

.31.

00-

0.59

136

95.1

7189

.91.

00-

0.90

AC

439

34.7

398

32.0

1.13

0.95

–1.3

37

4.9

78.

90.

620.

20–1

.91

CC

574.

559

4.7

0.99

0.68

–1.4

4-

-1

1.3

--

AC

+ C

C49

639

.245

736

.71.

110.

94–1

.30

74.

98

10.1

0.53

0.18

–1.5

8

BRC

A2rs

1801

406

AA

635

50.5

563

45.2

1.00

-0.

2279

54.5

4658

.31.

00-

0.53

AG

501

39.9

556

44.6

0.80

0.68

–0.9

458

40.0

3139

.21.

130.

62–2

.06

GG

121

9.6

127

10.2

0.84

0.64

–1.1

18

5.5

22.

53.

340.

65–1

7.12

AG

+ G

G62

249

.568

354

.80.

810.

69–0

.95

6645

.533

41.7

1.25

0.70

–2.2

5

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 15

Cau

casi

ans

Afr

ican

-Am

eric

an

Cas

esC

ontr

ols

Cas

esC

ontr

ols

Gen

eSN

PG

enot

ype

N%

N%

OR

†95

% C

IPe

rmut

ed P

*N

%N

%O

R†

95%

CI

Perm

uted

P*

BRC

A2rs

5433

04TT

821

65.1

839

67.6

1.00

-0.

6598

68.5

4455

.71.

00-

0.72

CT

399

31.6

364

29.3

1.12

0.94

–1.3

342

29.4

3038

.00.

590.

32–1

.10

CC

413.

338

3.1

1.11

0.70

–1.7

43

2.1

56.

30.

270.

06–1

.21

CT

+ C

C44

034

.940

232

.41.

120.

95–1

.32

4531

.535

44.3

0.54

0.30

–0.9

9

ERC

C2

(XPD

)rs

1052

555

CC

591

46.9

565

45.4

1.00

-0.

4711

076

.460

76.0

1.00

-0.

31

CT

549

43.6

535

43.0

0.98

0.83

–1.1

629

20.1

1721

.50.

820.

40–1

.68

TT11

99.

514

411

.60.

790.

60–1

.03

53.

52

2.5

1.56

0.27

–8.8

6

CT

+ TT

668

53.1

679

54.6

0.94

0.80

–1.1

034

23.6

1924

.00.

900.

45–1

.76

ERC

C2

(XPD

)rs

1318

1TT

505

41.0

480

39.1

1.00

-0.

1787

59.6

5060

.21.

00-

0.41

GT

575

46.6

571

46.5

0.96

0.81

–1.1

348

32.9

2833

.70.

850.

46–1

.57

GG

153

12.4

177

14.4

0.82

0.64

–1.0

511

7.5

56.

01.

190.

37–3

.78

GT

+ G

G72

859

.074

860

.90.

920.

79–1

.09

5940

.433

39.7

0.90

0.50

–1.6

1

ERC

C2

(XPD

)rs

1799

793

GG

545

44.0

527

43.2

1.00

-0.

0810

673

.665

79.3

1.00

-0.

77

AG

575

46.4

528

43.2

1.05

0.89

–1.2

531

21.5

1518

.31.

280.

62–2

.64

AA

120

9.7

166

13.6

0.70

0.54

–0.9

17

4.9

22.

42.

350.

44–1

2.43

AG

+ A

A69

556

.169

456

.80.

970.

83–1

.13

3826

.417

20.7

1.40

0.71

–2.7

7

ERC

C2

(XPD

)rs

2384

06G

G36

529

.038

330

.91.

00-

0.24

112

77.8

6580

.21.

00-

0.14

GT

636

50.4

600

48.5

1.11

0.93

–1.3

329

20.1

1619

.80.

980.

48–2

.00

TT26

020

.625

520

.61.

070.

86–1

.34

32.

10

0.0

--

GT

+ TT

896

71.1

855

69.1

1.10

0.93

–1.3

132

22.2

1619

.81.

090.

54–2

.20

ERC

C4

(XPF

)rs

1799

801

TT63

350

.362

350

.11.

00-

0.91

9666

.255

69.6

1.00

-0.

72

CT

521

41.4

521

41.9

0.98

0.84

–1.1

646

31.7

2329

.11.

270.

67–2

.40

CC

105

8.3

998.

01.

040.

78–1

.40

32.

11

1.3

3.02

0.29

–30.

9

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 16

Cau

casi

ans

Afr

ican

-Am

eric

an

Cas

esC

ontr

ols

Cas

esC

ontr

ols

Gen

eSN

PG

enot

ype

N%

N%

OR

†95

% C

IPe

rmut

ed P

*N

%N

%O

R†

95%

CI

Perm

uted

P*

CT

+ C

C62

649

.762

049

.90.

990.

85–1

.16

4933

.824

30.4

1.33

0.71

–2.4

9

ERC

C4

(XPF

)rs

1800

067

GG

1025

84.0

1012

83.0

1.00

-0.

7913

694

.478

96.3

1.00

-0.

59

AG

183

15.0

202

16.6

0.89

0.72

–1.1

18

5.6

33.

72.

120.

52–8

.54

AA

131.

15

0.4

2.58

0.92

–7.2

5-

--

--

-

AG

+ A

A19

616

.120

717

.00.

930.

75–1

.16

MG

MT

rs12

917

CC

949

75.9

916

74.1

1.00

-0.

8710

672

.160

74.1

1.00

-0.

68

CT

269

21.5

298

24.1

0.87

0.72

–1.0

535

23.8

2024

.71.

180.

61–2

.29

TT32

2.6

231.

91.

340.

78–2

.31

64.

11

1.2

2.75

0.30

–25.

41

CT

+ TT

301

24.1

321

26.0

0.91

0.75

–1.0

941

27.9

2125

.91.

260.

66–2

.40

MG

MT

rs23

0832

1A

A92

676

.792

277

.21.

00-

0.50

130

92.9

7391

.31.

00-

0.39

AG

267

22.1

256

21.4

1.04

0.86

–1.2

610

7.1

78.

80.

820.

29–2

.32

GG

141.

217

1.4

0.82

0.40

–1.6

7-

--

--

-

AG

+ G

G28

123

.327

322

.81.

020.

85–1

.24

MG

MT

rs23

0832

7A

A95

076

.196

077

.11.

00-

0.75

134

92.4

7590

.41.

00-

0.31

AG

276

22.1

266

21.4

1.05

0.87

–1.2

711

7.6

89.

60.

740.

27–2

.00

GG

221.

820

1.6

1.11

0.60

–2.0

5-

--

--

-

AG

+ G

G29

823

.928

623

.01.

050.

87–1

.26

MU

TYH

rs32

1948

4G

G10

8586

.410

7085

.71.

00-

0.77

140

97.9

7594

.91.

00-

0.77

AG

164

13.1

173

13.9

0.94

0.74

–1.1

83

2.1

45.

10.

290.

06–1

.36

AA

70.

65

0.4

1.39

0.44

–4.4

1-

--

--

-

AG

+ A

A17

113

.617

814

.30.

950.

76–1

.19

MU

TYH

rs94

2907

2A

A72

057

.168

655

.01.

00-

0.37

3423

.524

30.4

1.00

-0.

54

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 17

Cau

casi

ans

Afr

ican

-Am

eric

an

Cas

esC

ontr

ols

Cas

esC

ontr

ols

Gen

eSN

PG

enot

ype

N%

N%

OR

†95

% C

IPe

rmut

ed P

*N

%N

%O

R†

95%

CI

Perm

uted

P*

AG

455

36.1

485

38.9

0.89

0.76

–1.0

571

49.0

3240

.51.

390.

69–2

.79

GG

856.

876

6.1

1.07

0.77

–1.4

840

27.6

2329

.11.

110.

51–2

.39

AG

+ G

G54

042

.956

145

.00.

920.

78–1

.07

111

76.6

5569

.61.

270.

67–2

.42

OG

G1

rs32

1899

7C

C11

2691

.411

3791

.51.

00-

0.84

139

95.9

7797

.51.

00-

0.52

CT

106

8.6

106

8.5

1.01

0.76

–1.3

46

4.1

22.

51.

820.

34–9

.78

XPC

rs11

2654

7C

C96

976

.893

875

.41.

00-

0.33

109

75.7

6177

.21.

00-

0.53

CG

273

21.6

287

23.1

0.92

0.76

–1.1

130

20.8

1721

.50.

930.

46–1

.89

GG

201.

619

1.5

1.01

0.54

–1.9

15

3.5

11.

31.

920.

21–1

7.34

CG

+ G

G29

323

.230

624

.60.

930.

77–1

.11

3524

.318

22.8

0.99

0.50

–1.9

7

XPC

rs22

2800

1A

A45

736

.446

136

.91.

00-

0.66

7047

.636

43.4

1.00

-0.

23

AC

595

47.3

600

48.0

1.00

0.84

–1.1

961

41.5

3845

.80.

800.

44–1

.46

CC

205

16.3

190

15.2

1.09

0.86

–1.3

816

10.9

910

.81.

040.

40–2

.69

AC

+ C

C80

063

.679

063

.21.

020.

87–1

.20

7752

.447

56.6

0.85

0.48

–1.4

9

XPC

rs27

3353

2C

C46

136

.546

737

.61.

00-

0.74

7250

.039

47.6

1.00

-0.

41

CT

602

47.7

599

48.2

1.02

0.86

–1.2

157

39.6

3542

.70.

910.

50–1

.65

TT19

915

.817

614

.21.

140.

90–1

.46

1510

.48

9.8

1.19

0.45

–3.2

0

CT

+ TT

801

63.5

775

62.4

1.05

0.89

–1.2

372

50.0

4352

.40.

960.

54–1

.69

XPC

rs27

3353

4G

G34

027

.631

125

.41.

00-

0.39

5840

.039

47.6

1.00

-0.

66

CG

607

49.2

635

51.8

0.88

0.72

–1.0

664

44.1

3542

.71.

240.

67–2

.28

CC

286

23.2

279

22.8

0.94

0.75

–1.1

823

15.9

89.

81.

880.

74–4

.79

CG

+ C

C89

372

.491

474

.60.

900.

75–1

.07

8760

.043

52.4

1.36

0.77

–2.4

2

XPC

rs85

16TT

771

61.1

753

60.3

1.00

-0.

8510

673

.666

83.5

1.00

-0.

76

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 18

Cau

casi

ans

Afr

ican

-Am

eric

an

Cas

esC

ontr

ols

Cas

esC

ontr

ols

Gen

eSN

PG

enot

ype

N%

N%

OR

†95

% C

IPe

rmut

ed P

*N

%N

%O

R†

95%

CI

Perm

uted

P*

CT

425

33.7

436

34.9

0.95

0.81

–1.1

338

26.4

1316

.51.

730.

83–3

.60

CC

665.

260

4.8

1.08

0.75

–1.5

5-

--

--

-

CT

+ C

C49

138

.949

639

.70.

970.

83–1

.14

XRC

C1

rs17

9978

2C

C10

9888

.110

7186

.71.

00-

0.23

131

89.7

7286

.81.

00-

0.60

CT

143

11.5

158

12.8

0.88

0.69

–1.1

215

10.3

910

.80.

970.

38–2

.46

TT5

0.4

60.

50.

820.

25–2

.68

--

22.

4-

-

CT

+ TT

148

11.9

164

13.3

0.88

0.69

–1.1

115

10.3

1113

.20.

770.

32–1

.86

XRC

C1

rs25

487

GG

522

41.5

481

38.8

1.00

-0.

7710

371

.553

64.6

1.00

-0.

83

AG

576

45.8

590

47.6

0.90

0.76

–1.0

737

25.7

2733

.00.

670.

36–1

.25

AA

159

12.7

169

13.6

0.87

0.68

–1.1

14

2.8

22.

40.

970.

16–5

.84

AG

+ A

A73

558

.575

961

.20.

890.

76–1

.05

4128

.529

35.4

0.69

0.38

–1.2

7

XRC

C1

rs25

489

GG

1120

90.0

1145

91.4

1.00

-0.

2613

793

.876

91.6

1.00

-0.

61

AG

121

9.7

106

8.5

1.17

0.87

–1.5

39

6.2

78.

40.

590.

20–1

.76

AA

30.

22

0.2

1.55

0.26

–9.2

7-

--

--

-

AG

+ A

A12

410

.010

88.

61.

170.

89–1

.54

XRC

C1

rs28

5450

9C

C79

863

.477

762

.81.

00-

0.66

9867

.656

70.0

1.00

-0.

21

AC

406

32.3

400

32.3

0.99

0.83

–1.1

739

26.9

2328

.81.

090.

57–2

.08

AA

554.

461

4.9

0.88

0.60

–1.2

88

5.5

11.

32.

940.

34–2

5.66

AC

+ A

A46

136

.646

137

.20.

970.

83–1

.14

4732

.424

30.0

1.19

0.64

–2.2

1

XRC

C1

rs91

5927

AA

400

31.8

409

32.8

1.00

-0.

7362

42.8

3038

.01.

00-

0.41

AG

622

49.4

618

49.6

1.03

0.86

–1.2

354

37.2

3848

.10.

790.

42–1

.50

GG

238

18.9

220

17.6

1.11

0.88

–1.3

929

20.0

1113

.91.

320.

56–3

.11

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 19

Cau

casi

ans

Afr

ican

-Am

eric

an

Cas

esC

ontr

ols

Cas

esC

ontr

ols

Gen

eSN

PG

enot

ype

N%

N%

OR

†95

% C

IPe

rmut

ed P

*N

%N

%O

R†

95%

CI

Perm

uted

P*

AG

+ G

G86

068

.383

867

.21.

050.

89–1

.24

8357

.249

62.0

0.92

0.51

–1.6

5

Tota

l num

ber o

f cas

es a

nd c

ontro

ls in

bot

h C

auca

sian

s and

Afr

ican

-Am

eric

ans v

ary

by S

NPs

due

to m

issi

ng g

enot

ype

data

.

† OR

s and

95%

CI a

re a

djus

ted

for a

ge a

t dia

gnos

is (c

ases

) and

age

at r

efer

ence

dat

e (c

ontro

ls).

* Perm

uted

P-v

alue

s are

for t

he c

o-do

min

ant m

odel

s (2

degr

ees o

f fre

edom

) and

pre

sent

the

adju

sted

P-v

alue

s for

mul

tiple

com

paris

ons u

sing

a p

erm

utat

ion

proc

edur

e (s

ee st

atis

tical

met

hods

for d

etai

ls).

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 20

Tabl

e 3

Ass

ocia

tions

bet

wee

n SN

Ps in

DN

A re

pair

gene

s and

clin

ical

feat

ures

of p

rost

ate

canc

er a

mon

g C

auca

sian

s

Gle

ason

Sco

reC

ontr

ols

Cas

es* :

Gle

ason

scor

e 2–

6 or

7 (3

+4)

Cas

es* :

Gle

ason

scor

e 7

(4+3

) or

8–10

p-va

lue

OR

1≠ O

R2

Gen

eSN

PG

enot

ype

N%

N%

OR

1†95

% C

IN

%O

R2†

95%

CI

BRC

A2rs

1448

48TT

654

52.6

562

52.5

1.00

-92

47.2

1.00

-

GT

500

40.2

416

38.9

0.99

0.83

–1.1

881

41.5

1.16

0.84

–1.6

00.

34

GG

897.

292

8.6

1.32

0.95

–1.8

322

11.3

1.83

1.09

–3.0

80.

22

BRC

A2rs

1801

406

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

–0.9

085

43.8

0.86

0.63

–1.1

80.

42

GG

127

10.2

110

10.4

0.87

0.65

–1.1

711

5.7

0.48

0.25

–0.9

30.

08

ERC

C2

rs13

181

TT48

039

.141

840

.41.

00-

8443

.51.

00-

GT

571

46.5

492

47.5

0.96

0.80

–1.1

582

42.5

0.80

0.57

–1.1

10.

28

GG

177

14.4

126

12.2

0.75

0.57

–0.9

927

14.0

0.80

0.50

–1.2

80.

81

ERC

C2

rs17

9979

3G

G52

743

.245

043

.11.

00-

9247

.71.

00-

AG

528

43.2

490

47.0

1.06

0.88

–1.2

784

43.5

0.88

0.64

–1.2

10.

26

AA

166

13.6

103

9.9

0.67

0.50

–0.8

917

8.8

0.54

0.31

–0.9

30.

45

Tum

or S

tage

Con

trol

sC

ases

* : L

ocal

ized

Sta

geC

ases

* : R

egio

nal/D

ista

nt S

tage

p-va

lue

OR

1 ≠O

R2

Gen

eSN

PG

enot

ype

N%

N%

OR

1†95

% C

IN

%O

R2†

95%

CI

BRC

A2rs

1448

48TT

654

52.6

512

51.6

1.00

-14

351

.81.

00-

GT

500

40.2

393

39.6

1.02

0.85

–1.2

310

538

.01.

000.

75–1

.32

0.85

GG

897.

288

8.9

1.37

0.99

–1.9

028

10.1

1.57

0.98

–2.5

10.

57

BRC

A2rs

1801

406

AA

563

45.2

495

50.5

1.00

-14

050

.51.

00-

AG

556

44.6

389

39.7

0.76

0.64

–0.9

211

240

.40.

790.

60–1

.04

0.84

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Agalliu et al. Page 21

Tum

or S

tage

Con

trol

sC

ases

* : L

ocal

ized

Sta

geC

ases

* : R

egio

nal/D

ista

nt S

tage

p-va

lue

OR

1 ≠O

R2

Gen

eSN

PG

enot

ype

N%

N%

OR

1†95

% C

IN

%O

R2†

95%

CI

GG

127

10.2

969.

80.

810.

60–1

.10

259.

10.

790.

49–1

.26

0.89

ERC

C2

rs13

181

TT48

039

.138

940

.41.

00-

116

43.1

1.00

-

GT

571

46.5

450

46.7

0.94

0.78

–1.1

412

546

.50.

890.

67–1

.19

0.74

GG

177

14.4

125

13.0

0.79

0.60

–1.0

428

10.4

0.63

0.40

–0.9

90.

33

ERC

C2

rs17

9979

3G

G52

743

.242

243

.51.

00-

123

45.6

1.00

-

AG

528

43.2

448

46.2

1.02

0.85

–1.2

312

747

.01.

030.

78–1

.36

0.96

AA

166

13.6

100

10.3

0.68

0.51

–0.9

120

7.4

0.50

0.30

–0.8

30.

24

* Tota

l num

ber o

f cas

es a

nd c

ontro

ls v

ary

by S

NPs

due

to m

issi

ng g

enot

ype

data

. Cas

es w

ith m

issi

ng d

ata

on G

leas

on sc

ore

(n=6

) wer

e ex

clud

ed fr

om re

spec

tive

anal

yses

.

† OR

s and

95%

CI a

re a

djus

ted

for a

ge a

nd p

rost

ate

canc

er sc

reen

ing

hist

ory

with

in th

e 5

year

s bef

ore

pros

tate

can

cer d

iagn

osis

or r

efer

ence

dat

e.

Cancer Causes Control. Author manuscript; available in PMC 2011 February 1.