Predictive Markers and Cancer Prevention
Polymorphisms of DNA repair genes XRCC1 and XPD and their associations with risk of esophageal squamous cell carcinoma in a Chinese population
Article first published online: 20 JUN 2002
Copyright © 2002 Wiley-Liss, Inc.
International Journal of Cancer
Volume 100, Issue 5, pages 600–605, 10 August 2002
How to Cite
Xing, D., Qi, J., Miao, X., Lu, W., Tan, W. and Lin, D. (2002), Polymorphisms of DNA repair genes XRCC1 and XPD and their associations with risk of esophageal squamous cell carcinoma in a Chinese population. Int. J. Cancer, 100: 600–605. doi: 10.1002/ijc.10528
- Issue published online: 17 JUL 2002
- Article first published online: 20 JUN 2002
- Manuscript Accepted: 6 MAY 2002
- Manuscript Revised: 26 APR 2002
- Manuscript Received: 14 MAR 2002
- National Natural Science Foundation. Grant Numbers: 39825122, 39990570
- State Key Basic Research Program. Grant Number: G1998051200
- esophageal cancer;
- genetic polymorphism;
- DNA repair
Esophageal squamous cell carcinoma (ESCC), which is prevalent in China, is believed to be induced by environmental carcinogens. Accumulating evidence has shown that individual variation in DNA repair capacity resulting from genetic polymorphism influences risk of environmental carcinogenesis. We therefore investigated the associations between genetic polymorphisms in the DNA repair genes XRCC1 (Arg194Trp and Arg399Gln) and XPD (Asp312Asn and Lys751Gln) and risk of ESCC in an at-risk Chinese population. Genotypes were determined by a PCR-based approach in 433 patients with ESCC and 524 frequency-matched normal controls. We found that individuals with Trp/Trp genotype at XRCC1 Arg194Trp site had a 2-fold increased risk of this disease compared to Arg/Arg genotype (adjusted OR = 1.98; 95% CI 1.26–3.12). Furthermore, when compared to Arg/Arg and Arg/Trp genotype combined, homozygote for Trp/Trp genotype significantly increased the risk of developing ESCC, with the adjusted OR being 2.07 (95% CI 1.34–3.20). However, the XRCC1 Arg399Gln polymorphism was not significantly associated with risk of ESCC, with the adjusted OR being 0.87 (95% CI 0.55–1.37). Neither Asp312Asn nor Lys751Gln polymorphisms in the XPD gene influenced risk of ESCC in our study. These findings suggest that DNA repair gene XRCC1 but not XPD might play a role in esophageal carcinogenesis and might represent a genetic determinant in the development of the cancer. © 2002 Wiley-Liss, Inc.
It has been shown that reduction in DNA repair capacity is associated with increased risk for certain cancers. A wide diversity of DNA damage could be induced by normal metabolic processes in endogenous origin or by environmental carcinogens in exogenous sources. If not repaired, such damage can be converted into gene mutations and genomic instability, which in turn result in cellular malignant transformation. Nevertheless, cells have evolved surveillance mechanisms that monitor the integrity of genomes to minimize the consequences of detrimental mutations. The normal function of DNA repair enzymes is of importance for the damage-removing mechanisms. One polypeptide XRCC1 encoded by X-ray repair cross complementary 1 gene is implicated in the processes in single-strand break repair or base excision repair (BER).1, 2 Mutant hamster ovary cell lines that lack XRCC1 protein are hypersensitive to ionizing radiation, hydrogen peroxide and alkylating agents, which induce a 10-fold increase in the frequency of chromosome aberrations compared to the parental cells.1 The importance of XRCC1 to genetic stability is further indicated by that mice lacking XRCC1 die in early embryogenesis.3 The XRCC1 protein has been found as a molecular scaffold interacted with poly (ADP-ribose) polymerase, DNA polymerase-β and DNA ligase IIIα.2, 4, 5 Recently, XRCC1 has been found to stimulate the DNA kinase and phosphatase activities of polynucleotide kinase at damaged DNA termini and accelerate DNA single-strand break repair.6 Another important component, XPD (xeroderma pigmentosum complementary group D), an evolutionarily conserved ATP-dependent helicase involved in nucleotide excision repair (NER) pathway, has 2 functions: NER and basal transcription as part of the transcription factor complex, TFIIH.7 Mutations at different sites in XPD result in different clinical syndromes: xeroderma pigmentosum, Cockayne's syndrome and trichotiodystrophy. Xeroderma pigmentosum has been associated with a >1,000-fold increased risk of sunlight-induced skin cancer.8–10
Single nucleotide polymorphisms in several DNA repair genes that lead to amino acid substitution have been described, including polymorphisms of Arg194Trp and Arg399Gln in XRCC1 gene and Asp312Asn and Lys751Gln in XPD gene.11 Although biochemic and biologic characteristics of the variants are not determined, several phenotype analysis studies have been conducted to evaluate the functional relevance. Lunn et al.12 found that the XRCC1 399Gln allele was significantly associated with higher levels of both aflatoxin B1-DNA adducts and glycophorin A variants. Abdel-Rahman et al.13 reported that individuals with XRCC1 399Gln variant exhibited an increased sister chromatid exchange after treatment with tobacco-specific carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK). Similar results have also been found for smokers.14 Associations between XRCC1 polymorphisms and individual susceptibility to some types of cancer have been examined in case-control studies with the results being contradictory.15–23 Phenotype analysis and molecular epidemiologic studies on XPD genetic polymorphisms, Asp312Asn and Lys751Gln, have also been conducted by different groups.24–32
Esophageal squamous cell carcinoma (ESCC) is one of the most prevalent cancers in China, with an estimate of more than 250,000 cases diagnosed every year.33 Although the integrated etiology of ESCC remains to be fully elucidated, it has been shown that nitrosamine carcinogens, nutritional deficiency and other environmental risk factors may be major causes of this disease.34 However, only a fraction of exposed individuals develop ESCC, suggesting that individual genetic traits may influence the differential predisposition to this disease. Accumulating evidence has suggested that genetic variations in sensitivity to carcinogen exposure and DNA repair capacity might be important inherited risk components in carcinogenesis.35–38 To date, only one study has evaluated the role of the Arg194Trp and Arg399Gln polymorphisms in XRCC1 gene in relation to esophageal cancer in Taiwan,39 a region with relatively low incidence of the cancer (6.93/105 population40). To our best knowledge, no study has been published so far with regard to the association between the XPD polymorphisms and risk of ESCC. In view of the important role played by XRCC1 and XPD in maintaining genetic stability, which stand for 2 distinctive repair pathways, and possible effects on their protein function of the Arg194Trp and Arg399Gln polymorphisms in XRCC1 and Asp312Asn and Lys751Gln polymorphisms in XPD, we conducted a hospital-based, case-control study with a relatively large sample size in a Chinese population to analyze the contribution of these genetic polymorphisms to risk of ESCC.
MATERIAL AND METHODS
Our case-control study consisted of 433 patients with ESCC and 524 healthy controls. All subjects were unrelated ethnic Chinese. The patients with histologically confirmed primary ESCC were recruited without selection from January 1997–July 2001 in the Cancer Hospital of Chinese Academy of Medical Sciences (Beijing, China), yielding a 100% participation rate. All cases were from Beijing city and surrounding regions. Because this was a study of genotype and the marker was a constitutional one, both incident and prevalent cases were included. Some cases were subjects in a molecular epidemiologic study of susceptibility markers for ESCC as described previously.41, 42 In our current study, we extended the sample size of ESCC patients to 433. Population controls were randomly selected from a nutritional survey database conducted in the same regions during the time period of case collection. The characteristics of most control subjects were described previously.42 The criteria for the selection of controls included no history of cancer and frequency matching to cases on sex, age (± 5 years) and ethnicity. In our present study, we extended the sample size of controls to 524, which were derived from the same database. Personal data from each participant regarding demographic characteristics (e.g., sex and age) and related risk factors including tobacco smoking and family history of cancer were collected by questionnaire with the response rate being 100%. This study was approved by the Institutional Review Board of the Chinese Academy of Medical Sciences Cancer Institute.
Genomic DNA was extracted from peripheral blood of the controls or from surgically resected “normal” tissues adjacent to the tumor of ESCC patients, using standard procedures as described previously.43 Genotypes were analyzed using PCR-based methods as described below. Genotyping was accomplished with blinding to case/control status. A 15% masked, random sample of cases (n = 65) and controls (n = 79) was tested twice by different persons and the results were concordant for all masked duplicate sets.
The genotypes of XRCC1 at the Arg194Trp and Arg399Gln sites were determined by PCR-based restriction fragment length polymorphism (RFLP) methods as described by Sturgis et al.16 with some modifications. The primers for XRCC1 gene exon 6 containing Arg194Trp were: 5′ GCC AGG GCC CCT CCT TCA A 3′ and 5′ TAC CCT CAG ACC CAC GAG T 3′, which produce a 485 bp fragment. The primers for XRCC1 Arg399Gln polymorphism were: 5′ TCC TCC ACC TTG TGC TTT CT 3′ and 5′ AGT AGT CTG CTG GCT CTG GG 3′, which produce a 517 bp fragment. PCR was performed with a 25 μl reaction mixture containing approx. 100 ng DNA, 1.0 μM each primer, 0.2 mM each dNTP, 2.0 mM MgCl2, 1.0 U Taq DNA polymerase with 1 × reaction buffer (Promega, Madison, WI) and 2% DMSO. The reaction for amplification of the Arg194Trp site was carried out in the following conditions: an initial melting step of 2 min at 95°C, followed by 35 cycles of 30 sec at 94°C, 30 sec at 57°C and 45 sec at 72°C and a final elongation of 7 min at 72°C. The reaction conditions for the Arg399Gln site shared the same profile above-mentioned, except for annealing temperature at 61°C.
The restriction enzyme PvuII and NciI (New England Biolabs, Beverly, MA) were used to distinguish the Arg194Trp polymorphism at codon 194 and Arg399Gln polymorphism at codon 399, respectively. The wild-type Arg194 allele produces a single band representing the entire 485 bp fragments, and the variant 194Trp allele has 396 and 89 bp fragments because it gains a PvuII site. The wild-type allele (Arg/Arg) at the Arg399Gln site generates 2 DNA bands (384 and 133 bp), the variant allele (Gln/Gln) has a single 517 bp fragment, and the heterozygote (Arg/Gln) displays all 3 bands (517, 384 and 133 bp). The restricted products were analyzed by electrophoresis in a 2% agarose gel containing ethidium bromide.
The XPD genotypes at the Asp312Asn and Lys751Gln were analyzed as described previously,24, 26 with the primers for the Asp312Asn polymorphism being 5′-CTG TTG GTG GGT GCC CGT ATC TGT TGG TCT-3′/5′-TAA TAT CGG GGC TCA CCC TGC AGC ACT TCC T-3′ and for the Lys751Gln polymorphism being 5′-GCC CGC TCT GGA TTA TAC G-3′/5′-CTA TCA TCT CCT GGC CCC C-3′. The restriction enzyme StyI (New England Biolabs) was used to genotype the Asp312Asn polymorphism. The wild-type alleles (Asp/Asp) display 2 DNA bands (507 and 244 bp), the variant-type alleles (Asn/Asn) result in 3 bands (474, 244 and 33 bp), and the heterozygotes (Asp/Asn) have all 4 bands (507, 474, 244 and 33 bp). The restriction enzyme PstI (New England Biolabs) was used to distinguish the Lys751Gln polymorphism. PstI digestion resulted in 2 fragments of 290 and 146 bp for the wild-type alleles (Lys/Lys), 3 fragments (227, 146 and 63 bp) for the variant alleles (Gln/Gln) and 4 fragments (290, 227, 146 and 63 bp) for the heterozygotes (Lys/Gln). The 312Asp and 751Lys alleles are in linkage disequilibrium, that is, carriers with the XPD 312Asp allele tend to have the 751Lys allele.24, 27, 31 Therefore, we reanalyzed the 74 samples with the 312Asp allele but not 751Lys and with 312Asn but not 751Gln to ensure the genotyping. All of the results were compatible with the first analysis.
Chi (χ2) tests were used to examine the differences in the distributions of genotypes between cases and controls. The association between the XRCC1 or XPD polymorphisms and risk of ESCC was estimated by ORs and their 95% CIs, which were calculated by unconditional logistic regression. Smokers were considered current smokers if they smoked up to 1 year before the date of cancer diagnosis (or up to the date of the interview for controls). Information was collected on the number of cigarettes smoked per day, the age at which the subjects started smoking and the age at which ex-smokers stopped smoking. Pack-years smoked was calculated to indicate cumulative cigarette dose [pack-years = (cigarettes per day/20) × (years smoked)]. Because of the use of frequency matching, the ORs were also adjusted for age, gender and smoking status. All analyses were performed using Statistical Analysis System (Ver. 6.12, SAS Institute, Cary, NC).
We recruited 433 cases of ESCC and 524 normal controls in our study. All the subjects were Han Chinese. The distributions of age, sex and smoking status among cases and controls are summarized in Table I. There were no significant differences between cases and controls in terms of distributions of mean age, sex and smoking status, suggesting that the frequency matching was adequate. No special history of ESCC among cases and controls were reported.
|Variable||ESCC cases (n = 433)||Controls (n = 524)||p-value1|
|Male||333 (76.9)||390 (74.4)|
|Female||100 (23.1)||134 (25.6)|
|≤55||196 (45.3)||238 (45.4)|
|>55||237 (54.7)||286 (54.6)|
|Mean age in years (SD)||57.1 (9.73)||56.9 (7.22)||0.722|
|Smoking status, (%)||0.30|
|Never||183 (42.3)||239 (45.6)|
|Ever3||250 (57.7)||285 (54.4)|
Genotyping results (Table II) show that the allele frequencies for XRCC1 194Arg and 194Trp were 71.2% and 28.8% among controls compared to 67.6% and 32.4% among cases. The distribution of XRCC1 Arg194Trp genotypes among controls (Arg/Arg, 49.4%; Arg/Trp, 43.5%; Trp/Trp, 7.1%) did not deviate from that expected from the Hardy-Weinberg equilibrium (p = 0.59). The frequencies of the 3 genotypes of XRCC1 Arg194Trp among ESCC cases were Arg/Arg, 48.7%; Arg/Trp, 37.9%; and Trp/Trp, 13.4%, which were significantly different from those among controls (χ2 = 11.44, p = 0.003). The distribution of Arg/Arg, Arg/Gln and Gln/Gln genotypes at the Arg399Gln site among controls was 53.2%, 37.4% and 9.4%, which was also in accordance with the Hardy-Weinberg equilibrium (p = 0.48) but not significantly different from that among ESCC cases (58.0%, 33.9% and 8.1%, respectively) (Table II). Consistent with that reported by Shen et al.,18 our results also demonstrated that the 194 Trp/Trp genotype was exclusively linked to the 399 Arg/Arg genotype while the 399 Gln/Gln to 194 Arg/Arg both in cases and controls. Genotyping data of the XPD polymorphisms among cases and controls are also summarized in Table II. The frequencies of XPD Asp312Asn and Lys751Gln genotypes were in accordance with the Hardy-Weinberg equilibrium among the controls (p = 0.83 and p = 1.00, respectively).
|XRCC1Arg194Trp, n (%)||XRCC1 Arg399Gln, n (%)||XPDAsp312Asn, n (%)||XPD Lys751Gln, n (%)|
|Control||259 (49.4)||228 (43.5)||37 (7.1)||279 (53.2)||196 (37.4)||49 (9.4)||461 (88.0)||62 (11.8)||1 (0.2)||451 (86.1)||70 (13.3)||3 (0.6)|
|ESCC||211 (48.7)||164 (37.9)||58 (13.4)||0.003||251 (58.0)||147 (33.9)||35 (8.1)||0.34||381 (88.0)||49 (11.3)||3 (0.7)||0.48||367 (84.8)||63 (14.5)||3 (0.7)||0.84|
Individuals with the Trp/Trp genotype at the XRCC1 Arg194Trp site had a 2-fold increased risk of developing ESCC compared to those having the Arg/Arg genotype (adjusted OR = 1.98; 95% CI 1.26–3.12). Furthermore, when compared to Arg/Arg and Arg/Trp genotypes combined, homozygotes for Trp/Trp genotype were also associated with a significantly increased risk of ESCC, with the adjusted OR being 2.07 (95% CI 1.34–3.20). The increased risk of ESCC related to the polymorphism was not influenced by age, sex and smoking status. The XRCC1 Arg399Gln polymorphism was found not to be a risk-modulating factor for ESCC in our study (Table III). We also analyzed the potential interactions between the Arg194Trp and Arg399Gln polymorphisms and between the Arg194Trp polymorphism and smoking on risk of ESCC but did not observe any evidence for the effect (data not shown).
|Genotype||ESCC cases, n (%)||Controls, n (%)||Crude OR (95% CI)||Adjusted OR (95% CI)1|
|Arg/Arg||211 (48.7)||259 (49.4)||1.00||1.00|
|Arg/Trp||164 (37.9)||228 (43.5)||0.88 (0.67–1.16)||0.88 (0.67–1.15)|
|Trp/Trp||58 (13.4)||37 (7.1)||1.92 (1.23–3.02)||1.98 (1.26–3.12)|
|Arg/Arg or Arg/Trp||375 (86.6)||487 (92.9)||1.00||1.00|
|Trp/Trp||58 (13.4)||37 (7.1)||2.04 (1.32–3.14)||2.07 (1.34–3.20)|
|Arg/Arg||251 (58.0)||279 (53.2)||1.00||1.00|
|Arg/Gln||147 (33.9)||196 (37.4)||0.83 (0.63–1.10)||0.83 (0.63–1.09)|
|Gln/Gln||35 (8.1)||49 (9.4)||0.79 (0.50–1.27)||0.81 (0.51–1.29)|
|Arg/Arg or Arg/Gln||398 (91.9)||475 (90.6)||1.00||1.00|
|Gln/Gln||35 (8.1)||49 (9.4)||0.85 (0.54–1.34)||0.87 (0.55–1.37)|
The risk of ESCC related to XPD genotypes was also investigated in our study (Table IV). The 2 polymorphisms at this gene locus were found to be not associated with ESCC. Individuals who carried at least 1 variant 312Asn allele did not have a significantly higher risk of ESCC compared to those who carried the Asp/Asp wild-type genotype (adjusted OR = 1.00; 95% CI 0.68–1.49), so did with the Lys751Gln polymorphism (adjusted OR = 1.11; 95% CI 0.77–1.60).
|Genotype||ESCC cases, n (%)||Controls, n (%)||Crude OR (95% CI)||Adjusted OR (95% CI)1|
|Asp/Asp||381 (88.0)||461 (88.0)||1.00||1.00|
|Asp/Asn or Asn/Asn||52 (12.0)||63 (12.0)||1.00 (0.68–1.48)||1.00 (0.68–1.49)|
|Lys/Lys||367 (84.8)||451 (86.1)||1.00||1.00|
|Lys/Gln or Gln/Gln||66 (15.2)||73 (13.9)||1.11 (0.78–1.59)||1.11 (0.77–1.60)|
In our study, we investigated whether genetic polymorphisms in XRCC1, a protein that plays a central role in BER and single-strand break repair, and in XPD, a helicase involved in NER and basal transcription, could have an impact on the risk of developing ESCC. We found that the XRCC1 194Trp allele was associated with an elevated risk for ESCC among the Chinese population. With regard to the XPD gene, neither Asp312Asn nor Lys751Gln polymorphisms influenced risk of ESCC in this study population. It has been well known that DNA repair is a very important mechanism in protection against gene mutation and cancer initiation. XRCC1 protein is exclusively required for DNA BER, strand-break repair and maintenance of genetic stability.1–3, 6 Although the exact biochemic effects of these 2 genetic polymorphisms in XRCC1 are not fully characterized, the amino acid substitutions caused by the Arg194Trp and Arg399Gln polymorphisms are located in the important regions of the protein (between the binding domains of DNA polymerase-β and PARP and the BRCT-1 region, which harbored the PARP binding domain, respectively). Therefore, one may expect that these amino acid substitutions would change the function of the protein. The association between XRCC1 polymorphisms and susceptibility to ESCC is thus biologically plausible. The other potential predictors of ESCC risk such as age, sex and smoking did not influence the observed effect of XRCC1, suggesting that this single nucleotide polymorphism may be a genetic determinant in developing ESCC.
Of 957 subjects genotyped in our study, all of the 95 subjects carrying the 194 Trp/Trp genotype had exclusively the 399 Arg/Arg genotype and all of the 84 subjects with the 399 Gln/Gln genotype had exclusively the 194 Arg/Arg genotype. Based on these results, we hypothesized that either 194Trp or 399Gln allele would result in reduced DNA repair capacity and both of them were thus risk alleles. If both variant alleles existed in one haplotype simultaneously, the detrimental biologic consequences might be lethal with evolutionary consideration. Consistent with this hypothesis, previous studies have demonstrated that 194Trp was a risk allele for the development of colorectal cancer,17 breast cancer44 and prostate cancer,45 which were concordant with our findings. However, although several phenotype analysis studies have associated 399Gln allele with high levels of DNA adduct formation and frequency of sister chromatid exchange,12–15 we did not observe any association between the XRCC1 Arg399Gln polymorphism and risk of ESCC in this Chinese population. In addition, conflict data also exist concerning the published studies. For instance, in the previous investigations, XRCC1 194Trp allele was found not to be associated with risk of certain cancer19, 27, 39 or to be a protective factor against cancer risk.16, 18, 21 The reason for the contradictory results in these studies is currently unknown, but several factors relevant to these polymorphisms such as different carcinogen exposures in different populations and different types of DNA damage in the initiation of different cancers might cause differential results. Moreover, inadequate study design such as nonrandom sampling, limited sample size and pitfalls of unknown confounders should also be considered.
ESCC is a complex disease that may be attributed to the integrated outcome of exposure to endogenous and/or exogenous carcinogens.34 Therefore, interindividual differences in metabolic activation and detoxification process and DNA repair capacity reflecting the acquired and inherent host status may influence risk of developing this cancer. We have previously shown that genetic polymorphism in CYP2E1, a gene encoding the enzyme that is responsible for the metabolic activation of small molecule carcinogens such as nitrosamines, is an important modifier of risk for ESCC.46 In addition, low intake or low plasma levels of micronutrients that have antioxidative activity have been associated with increased risk of ESCC,47, 48 suggesting that oxidative DNA damage could also contribute to esophageal carcinogenesis. The small chemical-DNA base modifications and oxidative DNA damage are mainly targeted by the BER system in which XRCC1 is an important component.49 Findings supporting the implication of small base modifications and oxidative damage in ESCC development also come from the studies on the O6-alkylguanine-DNA alkyltransferase gene50 and the hOGG1 gene.51 Taken together, these data along with our findings in our present study suggest that the XRCC1 gene may play a role in protecting against ESCC.
On the other hand, we failed to observe any association between the genetic polymorphisms in the XPD gene and ESCC risk in our study. Because these XPD polymorphisms have been linked to reduced repair capacity for removal of benzo[a]pyrene diol epoxide-induced damage and ultraviolet-induced photoproducts from DNA in a host-cell reactivation assay,24, 52 our results showing a lack of association between XPD polymorphisms and risk of ESCC could imply that the NER system, a major pathway for repair of bulky DNA damage, might not be involved in repair of DNA damage produced by the carcinogen(s) having the capability to initiate ESCC in this population. However, caution should be made in interpreting these findings since it is well known that more than 20 genes are involved in the NER pathway and each individual could have been a wild type of one gene but a variant of another. Given this consideration, further study with analysis of the NER phenotype may be warranted to test this speculation. Furthermore, since this was a hospital-based case-control study, selection bias and/or systematic error might occur when we analyzed environmental factors such as smoking. It would be important to confirm these findings in a population-based prospective study.
Our study included 524 cancer-free ethnic Han Chinese as controls. We can thus provide an estimate of the frequency of XRCC1 gene polymorphisms in this population in north China. We observed an allelic frequency of 28.8% for the XRCC1 194Trp allele, which was similar to that (194Trp allele, 30.0%) reported in a Chinese population by Lee et al.39 (p = 0.65) but slightly lower (p = 0.04) than that (194Trp allele, 34.6%) reported by Shen et al.18 Several possibilities such as the sampling region of the study subjects and the relatively small sample size in their study might have caused the discrepancy.
In summary, our results demonstrate an association between the XRCC1 194Trp allele and elevated risk for ESCC in a Chinese population. These findings support the hypothesis that genetic variations in DNA repair protein contribute to risk of ESCC. Functional studies on the XRCC1 polymorphisms are warranted to confirm our results.
- 8DNA Repair. New York: W.H. Freedman and Company, 1985..
- 40Department of Health, Republic of China. Cancer registry annual report. Taipei, Republic of China: Department of Health, 1996.
- 44Polymorphisms of DNA repair genes in breast cancer. Proc Am Assoc Cancer Res 2000; 41: 128., , , , .
- 45Genetic polymorphism of DNA repair in human prostate cancer risk. Proc Am Assoc Cancer Res 2000; 41: 596., , , , , , , , , .