Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

APEX Nuclease (Multifunctional DNA Repair Enzyme) 1 Gene Asp148Glu Polymorphism and Cancer Risk: A Meta-Analysis Involving 58 Articles and 48903 Participants

  • Dan Hu,

    Affiliations Department of Pathology, Fujian Provincial Tumor Hospital, Teaching Hospital of Fujian Medical University, Fuzhou, Fujian province, China, Fujian Provincial Key Laboratory of Translational Center Medicine, Fujian Provincial Tumor Hospital, Teaching Hospital of Fujian Medical University, Fuzhou, Fujian province, China

  • Xiandong Lin,

    Affiliations Department of Pathology, Fujian Provincial Tumor Hospital, Teaching Hospital of Fujian Medical University, Fuzhou, Fujian province, China, Fujian Provincial Key Laboratory of Translational Center Medicine, Fujian Provincial Tumor Hospital, Teaching Hospital of Fujian Medical University, Fuzhou, Fujian province, China

  • Hejun Zhang,

    Affiliations Department of Pathology, Fujian Provincial Tumor Hospital, Teaching Hospital of Fujian Medical University, Fuzhou, Fujian province, China, Fujian Provincial Key Laboratory of Translational Center Medicine, Fujian Provincial Tumor Hospital, Teaching Hospital of Fujian Medical University, Fuzhou, Fujian province, China

  • Xiongwei Zheng ,

    niuwenquan_shcn@163.com (WN); agu1960@126.com (XZ)

    Affiliations Department of Pathology, Fujian Provincial Tumor Hospital, Teaching Hospital of Fujian Medical University, Fuzhou, Fujian province, China, Fujian Provincial Key Laboratory of Translational Center Medicine, Fujian Provincial Tumor Hospital, Teaching Hospital of Fujian Medical University, Fuzhou, Fujian province, China

  • Wenquan Niu

    niuwenquan_shcn@163.com (WN); agu1960@126.com (XZ)

    Affiliation State Key Laboratory of Medical Genomics, Ruijin Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China

APEX Nuclease (Multifunctional DNA Repair Enzyme) 1 Gene Asp148Glu Polymorphism and Cancer Risk: A Meta-Analysis Involving 58 Articles and 48903 Participants

  • Dan Hu, 
  • Xiandong Lin, 
  • Hejun Zhang, 
  • Xiongwei Zheng, 
  • Wenquan Niu
PLOS
x

Abstract

Background

Polymorphisms in the APEX nuclease (multifunctional DNA repair enzyme) 1 gene (APEX1) may be involved in the carcinogenesis by affecting DNA repair. We aimed to summarize available data on the association of the APEX1 Asp148Glu (rs1130409) polymorphism with risk of multiple types of cancer via a meta-analysis.

Methods and Results

In total, 58 qualified articles including 22,398 cancer patients and 26,505 controls were analyzed, and the data were extracted independently by two investigators. Analyses of the full data set indicated a marginally significant association of the APEX1 Asp148Glu polymorphism with cancer risk under allelic (odds ratio (OR)=1.05; 95% confidence interval (95% CI): 0.99-1.11; P=0.071), dominant (OR=1.09; 95% CI: 1.01-1.17; P=0.028), and heterozygous genotypic (OR=1.08; 95% CI: 1.01-1.16; P=0.026) models, with significant heterogeneity and publication bias. In subgroup analyses by cancer type, with a Bonferroni corrected alpha of 0.05/6, significant association was observed for gastric cancer under both dominant (OR=1.74; 95% CI: 1.2-2.51; P=0.003) and heterozygous genotypic (OR=1.66; 95% CI: 1.2-2.31; P=0.002) models. In subgroup analysis by ethnicity, risk estimates were augmented in Caucasians, especially under dominant (OR=1.11; 95% CI: 1.0-1.24; P=0.049) and heterozygous genotypic (OR=1.11; 95% CI: 0.99-1.24; P=0.063) models. By study design, there were no significant differences between population-based and hospital-based studies. In subgroup analysis by sample size, risk estimates were remarkably overestimated in small studies, and no significance was reached in large studies except under the heterozygous genotypic model (OR=1.23; 95% CI: 1.06-1.43; P=0.006, significant at a Bonferroni corrected alpha of 0.05/2). By quality score, the risk estimates, albeit nonsignificant, were higher in low-quality studies than in high-quality studies. Further meta-regression analyses failed to identify any contributory confounders for the associated risk estimates.

Conclusions

Our findings suggest that APEX1 Asp148Glu polymorphism might be a genetic risk factor for the development of gastric cancer. Further investigations on large populations are warranted.

Introduction

Polymorphisms in the APEX nuclease (multifunctional DNA repair enzyme) 1 gene (APEX1) may be involved in the carcinogenesis by correcting DNA damage [1]. The APEX1 encodes the major apurinic/apyrimidinic endonuclease in human cells, and the loss of bases in apurinic/apyrimidinic sites can usually block the progress of the DNA replication apparatus and cause mutations. Therefore, the genetic defects responsible for the repair capacity of the APEX1 are often regarded as the logical candidates for its functional investigations. It is worth noting that a single transition of the 1349th base pair T allele to G allele, inducing the substitution of the 148th amino acid aspartate (Asp) to glutamate (Glu) (Asp148Glu, rs1130409), in the 5th exon of the APEX1, has been extensively investigated in association with a wide range of cancers, such as lung cancer, breast cancer, and bladder cancer [2-4]. The results of individual association studies in the literature, however, are often controversial and inconclusive. Taking lung cancer as an example, the APEX1 148Glu allele was a risk-conferring factor in Caucasians [5], but a risk-reducing factor in Asians [6]. As a caveat, this lack of consistency might be attributable to the presence of genetic heterogeneity across ethnic populations, the insufficient sample sizes involved, and the possibly uncontrolled confounding effects. To shed some light on these issues and to generate more information, we sought to summarize available data on the association of the APEX1 Asp148Glu polymorphism with all types of cancers from both English and Chinese literature via a meta-analysis, and further to explore the potential sources of between-study heterogeneity and the possible existence of publication bias.

Methods

Meta-analysis of observational studies poses particular challenges owing to its inherent biases and divergences in study design. We therefore carried out this meta-analysis according to the guidelines set forth by the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) statement [7] (Please see the Checklist S1).

Search strategy

Four databases including the PubMed, EMBASE (Excerpta Medica database), Wanfang (http://www.wanfangdata.com.cn), and CNKI (China National Knowledge Infrastructure, http://www.cnki.net) were searched on May 1, 2013 for observational studies investigating the association between the APEX1 Asp148Glu polymorphism and all types of cancers. Subject terms used for the search were: ‘apurinic/apyrimidinic’, ‘APE1’, ‘APEX1’, ‘cancer’, ‘tumor’, ‘neoplasm’, combined with ‘gene’, ‘polymorphism’, ‘variant’, ‘mutation’, ‘allele’, or ‘genotype’. The reference lists of all the retrieved articles as well as those of reviews on the same topic were also searched to identify the additional missing articles. Searching results were limited to studies with a case-control design and articles published in the English or Chinese language.

Study selection

Two investigators (Dan Hu and Wenquan Niu) independently obtained the full texts of potentially eligible articles on the basis of their titles and abstracts. To avoid the double counting of the participants recruited in more than one publication, article authors were emailed for inquiry when necessary. In case of more than one publication from the same study population, the data from the most recent or the most complete publication were extracted.

Inclusion/exclusion criteria

Our analyses were limited to the studies that strictly fulfilled the following inclusion criteria (all points must be satisfied for inclusion): (1) clinical endpoint (dependent variable): all types of cancers; (2) study design: either retrospective or nested case-control design; (3) independent variables: the genotype and/or allele counts of the APEX1 Asp148Glu polymorphism. Studies were excluded (one point was sufficient for exclusion) if they investigated the progression, severity, phenotype modification, and the response to treatment or survival, as well as if they were conference abstracts, case reports or series, editorials, narrative reviews, and the non-English and non-Chinese articles.

Data extraction

The data were extracted from all the qualified articles independently by two investigators (Dan Hu and Wenquan Niu) according to a standardized Excel template (Microsoft Corp, Redmond, WA). The discrepancies were resolved by the discussion and review of original articles, and a consensus was reached finally.

The data were collected on the first author, year of publication, ethnicity of the study population, cancer type, study design, case-control status, the genotypes/alleles of the APEX1 Asp148Glu polymorphism between patients and controls, and the demographic data, if available, including age, gender, smoking, and drinking.

Quality score assessment

The study quality was evaluated by using a quality assessment score developed for genetic association studies by Thakkinstian and colleagues [8]. Total scores range from 0 (the worst) to 12 (the best). The criteria for quality assessment of genetic associations between the APEX1 Asp148Glu polymorphism and cancer are described in the Table S1.

Statistical analyses

In this meta-analysis, four genetic models of inheritance were performed for APEX1 Asp148Glu polymorphism including allelic model (the 148Glu allele versus the 148Asp allele), dominant model (the 148Glu/148Glu genotype plus the 148Glu/Asp genotype versus the 148Asp/Asp genotype), homozygous (the 148Glu/148Glu genotype versus the 148Asp/Asp genotype) and heterozygous (the 148Glu/Asp genotype versus the 148Asp/Asp genotype) genotypic models.

The random-effects model using the DerSimonian & Laird method was employed to compute the weighted odds ratios (ORs) and the corresponding 95% confidence intervals (95% CIs). Heterogeneity between studies was evaluated by the χ2 test, and was quantified by the inconsistency index (I2) statistic, which ranges from 0% to 100% and is defined as the percentage of the observed between-study variability that is due to heterogeneity rather than chance.

Predefined subgroup analyses were performed a priori according to the cancer type, ethnicity of the study populations (Caucasian, Asian, African-American, or mixed), study design (population-based or hospital-based), the total sample size (<300 subjects or ≥300 subjects), and the quality score (score <7 or score ≥7). For a certain cancer, the data were presented and summarized if there were three or more independent studies that provided the genotype or allele counts of the Asp148Glu polymorphism between patients and controls.

Meta-regression analyses were performed to estimate the extent to which different study-level variables, including age, smoking, drinking, and quality score, explained the potential heterogeneity of pooled effect estimates of the APEX1 Asp148Glu polymorphism on cancer risk.

Besides the Egger’s test, publication bias was evaluated by the trim-and-fill method, which can estimate the number and outcomes of theoretically missing studies due to publication bias. P<0.05 was considered statistical significance, except for the I2 and Egger’s statistics, for which significance was defined as P<0.10 [9]. All statistical analyses were conducted by the STATA software (StataCorp, TX, version 11.2 for Windows).

Results

Eligible articles

A flow diagram schematizing the process of article selection with specific reasons is presented in Figure 1. In total, 413 potentially relevant articles were identified after the initial search, and 58 of them were deemed as eligible after applying further inclusion/exclusion criteria [3-6,10-63]. All qualified articles, including 52 articles written in English and 6 articles in Chinese [39,48,51,52,55,57], were published between the year 2003 and 2013. Because five articles provided data by ethnicity, two by cancer type, and two by the presence of menopause, there were 68 independent populations for comparisons in final analyses.

Study characteristics

The baseline characteristics of all qualified populations are shown in Table 1, and the genotype distributions and allele frequencies of the APEX1 Asp148Glu polymorphism between cancer patients and controls of all qualified populations are presented in the Table S2. Of 68 qualified populations, 14 were conducted for lung cancer, 10 for colorectal cancer, 9 for bladder cancer, 8 for breast cancer, 6 for prostate cancer, 4 for gastric cancer, 2 for pancreatic cancer, 2 for head and neck cancer, 2 for leukaemia cancer, and 1 for melanoma, biliary tract, cervical, esophageal, thyroid, hepatocellular, gioma, cervical, renal, endometrical carcinoma, and prostate cancers, respectively. The quality scores of all 68 populations ranged from 3 to 12, with a mean value of 6.9 (standard deviation: 1.92). Moreover, there were 30 populations involving Caucasians, 29 involving Asians, 4 involving African-Americans, and 5 involving the mixed populations. There were 27 populations conducted on a population-based design and 41 on a hospital-based design. 32 of 68 populations (47.1%) had the total sample size (the sum of patients and controls) equal to or greater than 300 participants in this meta-analysis.

First author (year)Quality scoreCancer typeEthnicityDesignSample sizeAge (years)
CasesControlsCasesControls
Misra RR et al (2003)5LungCaucasianPopulation3153156059
Popanda O et al (2004)7LungCaucasianHospital4634606155
Ito H et al (2004)9LungAsianHospital17844962.962.6
Chen L et al (2005)6ProstateAfrican-AmericansPopulation1241166459
Chen L et al (2005)6ProstateCaucasianPopulation2282196462
Shen M et al (2005)5LungAsianPopulation1191135555
Broberg K et al (2005)6BladderCaucasianPopulation631586969
Zienolddiny S et al (2006)9LungCaucasianPopulation3434136560
Zhang Y et al (2006) (Postmenopausal)7BreastCaucasianPopulation839679NANA
Zhang Y et al (2006) (Premenopausal)7BreastCaucasianPopulation587434NANA
Terry PD et al (2006)6BladderMixedHospital23921565.763.3
Moreno V et al (2006)10ColorectalCaucasianHospital359312NANA
Li C et al (2006)6MelanomaCaucasianHospital602603NANA
Li J et al (2006)6PancreaticMixedHospital384357NANA
Li C et al (2007)6Head and neckCaucasianHospital830854NANA
Huang M et al (2007)5BladderCaucasianHospital59659063.9462.77
Figueroa JD et al (2007)7BladderCaucasianHospital115011496665
De Ruyck K et al (2007)6LungCaucasianHospital1101106261
Berndt S et al (2007)11ColorectalMixedPopulation767773NANA
Berndt S et al (2007)11ColorectalCaucasianPopulation720725NANA
Chang JS et al (2008)5LungMixedPopulation11329965.8566.3
Chang JS et al (2008)5LungAfrican-AmericansPopulation25528063.5161.81
Zhu R et al (2008)5LeukaemiaAsianHospital105108NANA
Tse D et al (2008)8EsophagealCaucasianHospital3124546464
Smith TR et al (2008)7BreastCaucasianHospital33641657.458.7
Smith TR et al (2008)7BreastAfrican-AmericansHospital637857.458.7
Shekari M et al (2008)6CervicalAsianHospital13818048.5548.81
Pardini B et al (2008)7ColorectalCaucasianHospital53253258.557.4
Mitra AK et al (2008)5BreastAsianPopulation155235NANA
Kasahara M et al (2008)6ColorectalAsianHospital6812167.367.4
Huang WY et al (2008)7Biliary tractAsianPopulation411786NANA
Chiang FY et al (2008)7ThyroidAsianHospital28346945.343.9
Andrew AS et al (2008)8BladderCaucasianHospital10291281NANA
Sangrajrang S et al (2008) (Postmenopausal)9BreastAsianHospital2391804845.3
Sangrajrang S et al (2008) (Premenopausal)9BreastAsianHospital2682454845.3
Narter KF et al (2009)4BladderCaucasianHospital834563.4359.98
Lu J et al (2009)9LungAsianPopulation500517NANA
Lo YL et al (2009)7LungAsianHospital73073060.7760.8
Liu Y et al (2009)7GliomaCaucasianPopulation373365NANA
Gangwar R et al (2009)7BladderAsianHospital2062505957.8
Agachan B et al (2009)3LungCaucasianHospital986751.2648.81
Ji L et al (2009)4HepatocellularAsianHospital500507NANA
Ye CC et al (2010)6ColorectalAsianHospital12315860.9NA
Wang M et al (2010)6BladderAsianHospital23425363.562.9
Palli D et al (2010)9GastricCaucasianPopulation31454868.855.5
Osawa K et al (2010)6LungAsianHospital10412066.367.3
Jelonek K et al (2010)5ColorectalCaucasianHospital103153NANA
Jelonek K et al (2010)5Head and neckCaucasianHospital104110NANA
Jelonek K et al (2010)5BreastCaucasianHospital91412NANA
Brevik A et al (2010)5ColorectalCaucasianPopulation304359NANA
Canbay E et al (2010)7GastricCaucasianPopulation5024760.0752.8
Agalliu I et al (2010)9ProstateCaucasianPopulation13081266NANA
Agalliu I et al (2010)9ProstateAfrican-AmericansPopulation14985NANA
Wang MM et al (2010)6CervicalAsianHospital30630646.8446.04
Huang LZ et al (2011)6LeukaemiaAsianHospital415519NANA
Li Z et al (2011)10LungAsianHospital45544359.6858.39
Kuasne H et al (2011)4ProstateMixedHospital17217265.6463.86
Gu D et al (2011)7GastricAsianHospital33836261.7662.46
Cao Q et al (2011)6RenalAsianHospital61263256.956.7
Canbay E et al (2011)9ColorectalCaucasianPopulation7924760.2259.73
Deng Q et al (2011)4LungAsianPopulation3153155958
Zhonghua L et al (2011)5GastricAsianHospital12615658.753.1
Nakao M et al (2012)9PancreaticAsianPopulation1851465NANA
Mittal RD et al (2012)9ProstateAsianPopulation1952506664.7
Mittal RD et al (2012)9BladderAsianPopulation212250NANA
Mandal R et al (2012)12ProstateAsianPopulation19222462.659.1
Cincin Z et al (2012)4Endometrial carcinomaCaucasianHospital10415856.253.71
Li Y et al (2013)6ColorectalAsianHospital45163159.457

Table 1. The baseline characteristics of the study populations analyzed in this meta-analysis.

Abbreviations: NA, not available.
CSV
Download CSV

Overall analyses

Analyses of the full data set indicated a marginally significant association of the APEX1 Asp148Glu polymorphism with cancer risk under allelic (OR=1.05; 95% CI: 0.99-1.11; P=0.071), dominant (OR=1.09; 95% CI: 1.01-1.17; P=0.028), and heterozygous genotypic (OR=1.08; 95% CI: 1.01-1.16; P=0.026) models, with high probabilities of heterogeneity (I2=70.6%, 67.1%, and 59.5% respectively, all P<0.0005 from the χ2 test) (Table 2 and Table 3). Moreover, the probability of publication bias was high as reflected by both the Egger’s tests and the trim-and-fill funnel plots for these three models (Figure 2). We estimated that there were respectively 10, 11, and 10 missing independent populations to make the funnel plots symmetrical under allelic, dominant, and heterozygous genotypic models.

Groups/subgroupsNumber of studies (cases/controls)Allelic modelDominant model
OR; 95% CI; PI2 (P)PEggerOR; 95% CI; PI2 (P)PEgger
Overall estimates68 (22398/26505)1.05; 0.99-1.11; 0.07170.6% (<0.0005)0.0491.09; 1.01-1.17; 0.02867.1% (<0.0005)0.003
Cancer type
Lung cancer14 (4007/4513)1.06; 0.95-1.19; 0.32566.8% (<0.0005)0.0181.1; 0.93-1.3; 0.26867.6% (<0.0005)0.01
Colorectal cancer10 (3459/3978)1.07; 0.94-1.22; 0.32572.2% (<0.0005)0.8141.2; 0.97-1.49; 0.10175.2% (<0.0005)0.681
Bladder cancer9 (3618/3918)0.99; 0.92-1.06; 0.7013.4% (0.406)0.4810.99; 0.89-1.11; 0.90310.4% (0.348)0.058
Breast cancer8 (2546/2655)1.03; 0.88-1.21; 0.69569.3% (0.002)0.681.05; 0.82-1.34; 0.70471.8% (0.001)0.681
Prostate cancer6 (2122/2046)1.08; 0.98-1.2; 0.115.7% (0.38)0.1031.13; 0.95-1.35; 0.17228.9% (0.218)0.191
Gastric cancer4 (803/1311)1.42; 1.09-1.84; 0.00971.0% (0.016)0.161.74; 1.2-2.51; 0.00364.9% (0.036)0.082
Ethnicity
Caucasian30 (12044/13249)1.06; 0.99-1.13; 0.11666.5% (<0.0005)0.0221.11; 1.0-1.24; 0.04967.8% (<0.0005)0.011
Asian29 (8161/10945)1.03; 0.64-1.14; 0.50878.8% (<0.0005)0.6171.05; 0.93-1.19; 0.43871.6% (<0.0005)0.076
African-American4 (573/546)1.03; 0.86-1.22; 0.7620.0% (0.578)0.560.98; 0.77-1.25; 0.8680.0% (0.507)0.461
Mixed5 (1620/1765)1.07; 0.92-1.23; 0.37544.1% (0.128)0.6371.2; 0.95-1.53; 0.13254.2% (0.068)0.802
Study design
Population-based27 (8984/11489)1.04; 0.97-1.11; 0.25553.7% (0.001)0.0541.10; 0.99-1.22; 0.08560.9% (<0.0005)0.035
Hospital-based41 (13414/15016)1.05; 0.98-1.14; 0.18776.7% (<0.0005)0.251.08; 0.97-1.19; 0.14870.8% (<0.0005)0.039
Sample size
≥300 participants32 (17084/18154)0.99; 0.94-1.04; 0.66763.2% (<0.0005)0.0710.99; 0.93-1.06; 0.83450.2% (0.001)0.509
<300 participants36 (5314/8351)1.16; 1.05-1.3; 0.00673.5% (<0.0005)0.0161.26; 1.08-1.47; 0.00373.1% (<0.0005)0.003
Quality score
≥734 (13846/16752)1.03; 0.98-1.08; 0.23846.0% (0.0085)0.2021.06; 0.98-1.14; 0.15249.1% (0.001)0.061
<7(8477/9718)1.07; 0.97-1.19; 0.17580.7% (<0.0005)0.1431.13; 0.98-1.3; 0.09976.8% (<0.0005)0.019

Table 2. Overall and subgroup estimates of the associations of APEX1 Asp148Glu polymorphism with cancer risk under allelic and dominant models.

Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval.
CSV
Download CSV
Groups/subgroupsHomozygous genotypic modelHeterozygous genotypic model
OR; 95% CI; PI2 (P)PEggerOR; 95% CI; PI2 (P)PEgger
Overall estimates1.06; 0.96-1.17; 0.23662.5% (<0.0005)0.4891.08; 1.01-1.16; 0.02659.5% (<0.0005)0.002
Cancer type
Lung cancer1.07; 0.87-1.3; 0.53754.9% (0.009)0.0581.11; 0.93-1.32; 0.2665.9% (<0.0005)0.008
Colorectal cancer1.03; 0.8-1.33; 0.81565.1 % (0.005)0.1581.25; 1.0-1.56; 0.05574.7% (<0.0005)0.529
Bladder cancer0.94; 0.71-1.26; 0.68656.5% (0.032)0.4821.0; 0.9-1.11; 0.9743.3% (0.404)0.045
Breast cancer1.0; 0.78-1.27; 0.96743.9% (0.086)0.6871.05; 0.82-1.34; 0.69767.9% (0.003)0.703
Prostate cancer1.15; 0.95-1.4; 0.1480.0% (0.705)0.0011.1; 0.91-1.33; 0.59129.4% (0.214)0.271
Gastric cancer1.79; 1.11-2.89; 0.01764.2% (0.039)0.3321.66; 1.2-2.31; 0.00250.7% (0.107)0.054
Ethnicity
Caucasian1.06; 0.94-1.2; 0.33254.5% (<0.0005)0.2131.11; 0.99-1.24; 0.06365.1% (<0.0005)0.014
Asian1.04; 0.85-1.27; 0.72374.7% (<0.0005)0.6461.05; 0.94-1.17; 0.39658.1% (<0.0005)0.033
African-American1.11; 0.77-1.61; 0.5730.0% (0.71)0.5330.94; 0.73-1.22; 0.6460.0% (0.554)0.421
Mixed1.05; 0.81-1.36; 0.72421.2% (0.28)0.7081.24; 0.97-1.58; 0.08352.1% (0.08)0.83
Study design
Population-based1.03; 0.92-1.16; 0.57133.2% (0.052)0.1511.12; 1.0-1.26; 0.05163.2% (<0.0005)0.025
Hospital-based1.06; 0.92-1.23; 0.42671.9% (<0.0005)0.981.06; 0.97-1.16; 0.21557.1% (<0.0005)0.043
Sample size
≥300 participants1.21; 0.98-1.51; 0.08264.6% (<0.0005)0.1641.23; 1.06-1.43; 0.00669.1% (<0.0005)0.812
<300 participants0.99; 0.9-1.09; 0.84957.3% (<0.0005)0.9181.01; 0.95-1.07; 0.79731.1% (0.05)0.005
Quality score
≥71.05; 0.95-1.16; 0.31743.5% (0.005)0.7361.06; 0.98-1.15; 0.13150.8% (<0.0005)0.056
<71.08; 0.89-1.32; 0.43373.6% (<0.0005)0.5361.12; 0.98-1.27; 0.08766.6% (<0.0005)0.011

Table 3. Overall and subgroup estimates of the associations of APEX1 Asp148Glu polymorphism with cancer risk under two genotypic models.

Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval.
CSV
Download CSV
thumbnail
Figure 2. Trim-and-fill funnel plots for the effect of the APEX1 Asp148Glu polymorphism on cancer risk under four genetic models.

Hollow circles are the actual studies included in this meta-analysis, and solid squares are missing studies required to achieve symmetry.

https://doi.org/10.1371/journal.pone.0083527.g002

Subgroup analyses

To account for the potential sources of between-study heterogeneity, a set of predefined subgroup analyses were conducted (Table 2, Table 3, and Figures S1-S5).

By cancer type, after the Bonferroni correction for the multiple testing (Bonferroni significance threshold P=0.05 divided by the number of cancers (n=6): P=0.0083), significant association was observed for gastric cancer under both dominant (OR=1.74; 95% CI: 1.2-2.51; P=0.003) and heterozygous genotypic (OR=1.66; 95% CI: 1.2-2.31; P=0.002) models, whereas no significance was reached for the other cancers under investigation. The heterogeneity between studies was relatively low for bladder and prostate cancers.

By ethnicity, the magnitude of risk estimates was marginally significant in Caucasians under both dominant (OR=1.11; 95% CI: 1.0-1.24; P=0.049) and heterozygous genotypic (OR=1.11; 95% CI: 0.99-1.24; P=0.063) models, whereas this significance failed to survive the stringent Bonferroni correction (Bonferroni significance threshold P=0.05 divided by the number of ethnicities (n=4): P=0.0125). In Asians and African-Americans, there was no significant association observed in this meta-analysis.

By study design, there were no significant differences in the pooled risk estimates between the population-based and hospital-based studies, with high probabilities of between-study heterogeneity and publication bias.

By sample size, the risk estimates were significantly overestimated in small studies (the total sample size <300 participants), and no significance was reached in large studies (the total sample size ≥300 participants) under all but heterozygous genotypic model (OR=1.23; 95% CI: 1.06-1.43; P=0.006), even after the Bonferroni correction (Bonferroni significance threshold P=0.05 divided by the number of 2 groups: P=0.025). There was moderate evidence of heterogeneity.

By quality score, the risk estimates were relatively higher in low-quality studies (quality score <7) than in high-quality studies (quality score ≥7), and there was no significance observed under all four genetic models. The presence of heterogeneity was more evident in low-quality studies than in high-quality studies. Significant publication bias was found under both dominant and heterozygous genotypic models.

Meta-regression analyses

To further explore additional sources of between-study heterogeneity, we constructed a multivariable meta-regression model that included age, smoking, drinking, and quality score as independent variables. However, none of these variables were observed to significantly affect the relationship between the APEX1 Asp148Glu polymorphism and cancer susceptibility.

Discussion

Via a meta-analysis of the data from 58 articles and on 48903 participants, we investigated the association of the non-synonymous polymorphism Asp148Glu in APEX1 with cancer risk. The principle finding of this study was that the APEX1 148Glu allele was associated with the significant risk of developing gastric cancer under both dominant and heterozygous genotypic models, even after the Bonferroni correction. Moreover, our subgroup analyses indicated that ethnicity might be an underlying cause of heterogeneity between studies. Although other sources of heterogeneity cannot be easily ruled out, this study, to the best of our knowledge, is so far the largest meta-analysis examining the association of the APEX1 Asp148Glu polymorphism with cancer risk.

Recently, Zhou and colleagues have synthesized data from 32 case-control articles on the two polymorphisms of APEX1, and they failed to find any relationship between cancer risk and the Asp148Glu polymorphism [64]. By contrast, the findings of this meta-analysis supported the significant roles of the 148Glu allele in susceptibility to gastric cancer. However, a note of caution should be added because the risk estimates for gastric cancer were based on 803 patients and 1311 controls from 4 independent populations in this meta-analysis, the sample size might not be sufficient enough to derive a firm conclusion. It is recommended that to generate robust data, a much larger sample set encompassing more than 1000 participants in each group might be required [65]. A large, well-designed study is therefore warranted to confirm or refute the significance of our findings.

Moreover, extending the findings of the meta-analysis by Zhou and colleagues [64], we, in subgroup analyses, observed a marginally significant association of the APEX1 Asp148Glu polymorphism with cancer risk in Caucasians under both dominant and heterozygous genotypic models, but not in Asians and African-Americans. One possible explanation for this divergence is the genetic heterogeneity across ethnicities. For example in this meta-analysis, the average frequency of the APEX1 148Glu allele was 34.82% in Asian controls, but was as exceedingly high as 45.21% in Caucasian controls. In general, genetic heterogeneity is an inevitable problem in any disease identification strategy. This ethnicity-specific effect suggests that different genetic backgrounds may account for this discrepancy or that different populations may have different linkage disequilibrium patterns due to the evolutionary history. As such, it is necessary to construct a database of susceptible genes and polymorphisms implicated in carcinogenesis in each ethnic group.

To seek additional sources of heterogeneity, an alternative method is to perform a meta-regression analysis; however, none of the confounders under study contributed remarkably to the presence of heterogeneity in this meta-analysis. It is important to bear in mind that meta-regression analysis, albeit enabling quantitative covariates to be considered, does not have the methodological rigor of a properly designed study that is intended to test the effect of these covariates formally. Admittedly, one limitation facing this method was the number of available studies with detailed information such as smoking and drinking. In fact, most studies did not report the study-level covariates of interest, precluding a more robust assessment of additional sources of heterogeneity.

Some limitations need to be acknowledged for this meta-analysis. First, all qualified studies were conducted on case-control design, which precludes further comments on a cause-effect relationship. Second, in both overall and subgroup analyses, most resultant associations might be biased by the moderate to high degree of between-study heterogeneity, which enhances the difficulty in drawing firm conclusions and encourages the exploration of other possible reasons for heterogeneity. Third, the overall findings of this study were skewed by publication bias, although publication bias was improved in most subgroups, possibly due to the lack of power for small number of studies involved. Factually as suggested by Hannah and colleagues, the study power is low if the number of studies included in a meta-analysis is 10 or fewer [66]. Moreover, potential selection bias cannot be completely ruled out, because we only retrieved studies from English and Chinese journals and published articles. Fourth, due to the relatively small sample sizes involved in subgroup analyses, we must hold some reservations about the interpretation of our subgroup results. Last but not the least, we only focused on the APEX1 Asp148Glu polymorphism, and did not cover the other polymorphisms of APEX1. It is possible that the potential role of the examined polymorphism is diluted or masked by other gene-gene or gene-environment interactions. Thus, we cannot just to a conclusion until further confirmation of our findings has been undertaken.

In conclusion, via a meta-analysis of the data from 58 articles and on 48903 participants, we provide evidence that the APEX1 Asp148Glu polymorphism might be a genetic risk factor for the development of gastric cancer. Nevertheless, despite the small sample sizes involved in subgroup analyses, this meta-analysis provides an anchoring point for better understanding of the pathogenesis of cancers. For practical reasons, we hope that this study will not remain just another endpoint of research instead of a beginning to establish the background data to understand the roles of the APEX1 in carcinogenesis.

Supporting Information

Table S1.

Criteria for quality assessment of genetic associations of the APEX1 Asp148Glu polymorphism with cancer risk.

https://doi.org/10.1371/journal.pone.0083527.s001

(DOC)

Table S2.

The genotype distributions and allele frequencies of the APEX1 Asp148Glu polymorphism between cancer patients and controls of all examined populations in this meta-analysis.

https://doi.org/10.1371/journal.pone.0083527.s002

(DOC)

Figure S1.

Forest plots of the lung, bladder, colorectal cancer (the upper panel), and prostate, breast, gastric cancers (the lower panel) in subgroup analyses by cancer type for the APEX1 Asp148Glu polymorphism under the allelic model.

https://doi.org/10.1371/journal.pone.0083527.s003

(PDF)

Figure S2.

Forest plots of the Caucasians (the upper panel), Asians (the middle panel), African-Americans and mixed populations (the lower panel) in subgroup analyses by ethnicity for the APEX1 Asp148Glu polymorphism under the allelic model.

https://doi.org/10.1371/journal.pone.0083527.s004

(PDF)

Figure S3.

Forest plots of the hospital-based studies (the upper panel), and population-based studies (the lower panel) in subgroup analyses by study design for the APEX1 Asp148Glu polymorphism under the allelic model.

https://doi.org/10.1371/journal.pone.0083527.s005

(PDF)

Figure S4.

Forest plots of the small studies (the upper panel), and large studies (the lower panel) in subgroup analyses by sample size for the APEX1 Asp148Glu polymorphism under the allelic model.

https://doi.org/10.1371/journal.pone.0083527.s006

(PDF)

Figure S5.

Forest plots of the high-quality studies (the upper panel), and low-quality studies (the lower panel) in subgroup analyses by sample size for the APEX1 Asp148Glu polymorphism under the allelic model.

https://doi.org/10.1371/journal.pone.0083527.s007

(PDF)

Author Contributions

Conceived and designed the experiments: XZ WN. Performed the experiments: DH WN. Analyzed the data: XL HZ. Contributed reagents/materials/analysis tools: DH XL HZ. Wrote the manuscript: WN.

References

  1. 1. Raffoul JJ, Heydari AR, Hillman GG (2012) DNA Repair and Cancer Therapy: Targeting APE1/Ref-1 Using Dietary Agents. J Oncol 2012: 370481.
  2. 2. Lin CH, Chen PM, Cheng YW, Chen CY, Yuan CJ et al. (2012) The APE1 Asp/Asp genotype and the combination of APE1 Asp/Asp and hOGG1-Cys variants are associated with increased p53 mutation in non-small cell lung cancer. J Epidemiol 22: 537-542.
  3. 3. Canbay E, Cakmakoglu B, Zeybek U, Sozen S, Cacina C et al. (2011) Association of APE1 and hOGG1 polymorphisms with colorectal cancer risk in a Turkish population. Curr Med Res Opin 27: 1295-1302.
  4. 4. Wang M, Qin C, Zhu J, Yuan L, Fu G et al. (2010) Genetic variants of XRCC1, APE1, and ADPRT genes and risk of bladder cancer. DNA Cell Biol 29: 303-311.
  5. 5. De Ruyck K, Szaumkessel M, De Rudder I, Dehoorne A, Vral A et al. (2007) Polymorphisms in base-excision repair and nucleotide-excision repair genes in relation to lung cancer risk. Mutat Res 631: 101-110.
  6. 6. Deng Q, Sheng L, Su D, Zhang L, Liu P et al. (2010) Genetic polymorphisms in ATM, ERCC1, APE1 and iASPP genes and lung cancer risk in a population of southeast China. Med Oncol 28: 667-672.
  7. 7. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD et al. (2000) Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 283: 2008-2012.
  8. 8. Thakkinstian A, McEvoy M, Minelli C, Gibson P, Hancox B et al. (2005) Systematic review and meta-analysis of the association between {beta}2-adrenoceptor polymorphisms and asthma: a HuGE review. Am J Epidemiol 162: 201-211.
  9. 9. Bowden J, Tierney JF, Copas AJ, Burdett S (2011) Quantifying, displaying and accounting for heterogeneity in the meta-analysis of RCTs using standard and generalised Q statistics. BMC Med Res Methodol 11: 41.
  10. 10. Misra RR, Ratnasinghe D, Tangrea JA, Virtamo J, Andersen MR et al. (2003) Polymorphisms in the DNA repair genes XPD, XRCC1, XRCC3, and APE/ref-1, and the risk of lung cancer amongmale smokers in Finland. Cancer Lett 191: 171-178.
  11. 11. Ito H, Matsuo K, Hamajima N, Mitsudomi T, Sugiura T et al. (2004) Gene-environment interactions between the smoking habit and polymorphisms in the DNA repair genes, APE1 Asp148Glu and XRCC1 Arg399Gln, in Japanese lung cancer risk. Carcinogenesis 25: 1395-1401.
  12. 12. Popanda O, Schattenberg T, Phong CT, Butkiewicz D, Risch A et al. (2004) Specific combinations of DNA repair gene variants and increased risk for non-small cell lung cancer. Carcinogenesis 25: 2433-2441.
  13. 13. Broberg K, Bjork J, Paulsson K, Hoglund M, Albin M (2005) Constitutional short telomeres are strong genetic susceptibility markers for bladder cancer. Carcinogenesis 26: 1263-1271.
  14. 14. Shen M, Berndt SI, Rothman N, Mumford JL, He X et al. (2005) Polymorphisms in the DNA base excision repair genes APEX1 and XRCC1 and lung cancer risk in Xuan Wei, China. Anticancer Res 25: 537-542.
  15. 15. Chen L, Ambrosone CB, Lee J, Sellers TA, Pow-Sang J et al. (2006) Association Between Polymorphisms in the DNA Repair Genes XRCC1 and APE1, and the Risk of Prostate Cancer in White and Black Americans. J Urol 175: 108-112.
  16. 16. Li C, Liu Z, Wang LE, Strom SS, Lee JE et al. (2006) Genetic variants of the ADPRT, XRCC1 and APE1 genes and risk of cutaneous melanoma. Carcinogenesis 27: 1894-1901.
  17. 17. Moreno V, Gemignani F, Landi S, Gioia-Patricola L, Chabrier A et al. (2006) Polymorphisms in genes of nucleotide and base excision repair: risk and prognosis of colorectal cancer. Clin Cancer Res 12: 2101-2108.
  18. 18. Terry PD, Umbach DM, Taylor JA (2006) APE1 genotype and risk of bladder cancer: Evidence for effect modification by smoking. Int J Cancer 118: 3170-3173.
  19. 19. Zhang Y (2006) Genetic Polymorphisms in Base-Excision Repair Pathway Genes and Risk of Breast. Cancer Cancer Epidemiol Biomarkers Prev 15: 353-358.
  20. 20. Zienolddiny S, Campa D, Lind H, Ryberg D, Skaug V et al. (2006) Polymorphisms of DNA repair genes and risk of non-small cell lung cancer. Carcinogenesis 27: 560-567.
  21. 21. Berndt SI, Huang WY, Fallin MD, Helzlsouer KJ, Platz EA et al. (2007) Genetic Variation in Base Excision Repair Genes and the Prevalence of Advanced Colorectal Adenoma. Cancer Res 67: 1395-1404.
  22. 22. Figueroa JD, Malats N, Real FX, Silverman D, Kogevinas M et al. (2007) Genetic variation in the base excision repair pathway and bladder cancer risk. Hum Genet 121: 233-242.
  23. 23. Huang M, Dinney CP, Lin X, Lin J, Grossman HB et al. (2007) High-Order Interactions among Genetic Variants in DNA Base Excision Repair Pathway Genes and Smoking in Bladder Cancer Susceptibility. Cancer Epidemiol Biomarkers Prev 16: 84-91.
  24. 24. Li C, Hu Z, Lu J, Liu Z, Wang LE et al. (2007) Genetic polymorphisms in DNA base-excision repair genes ADPRT, XRCC1, and APE1 and the risk of squamous cell carcinoma of the head and neck. Cancer 110: 867-875.
  25. 25. Sangrajrang S, Schmezer P, Burkholder I, Waas P, Boffetta P et al. (2007) Polymorphisms in three base excision repair genes and breast cancer risk in Thai women. Breast Cancer Res Treat 111: 279-288.
  26. 26. Andrew AS, Karagas MR, Nelson HH, Guarrera S, Polidoro S et al. (2008) DNA Repair Polymorphisms Modify Bladder Cancer Risk: A Multi-factor Analytic Strategy. Hum Hered 65: 105-118.
  27. 27. Chang JS, Wrensch MR, Hansen HM, Sison JD, Aldrich MC et al. (2008) Base excision repair genes and risk of lung cancer among San Francisco Bay Area Latinos and African-Americans. Carcinogenesis 30: 78-87.
  28. 28. Chiang FY, Wu CW, Hsiao PJ, Kuo WR, Lee KW et al. (2008) Association between Polymorphisms in DNA Base Excision Repair Genes XRCC1, APE1, and ADPRT and Differentiated Thyroid Carcinoma. Clin Cancer Res 14: 5919-5924.
  29. 29. Huang WY, Gao YT, Rashid A, Sakoda LC, Deng J et al. (2008) Selected base excision repair gene polymorphisms and susceptibility to biliary tract cancer and biliary stones: a population-based case-control study in China. Carcinogenesis 29: 100-105.
  30. 30. Kasahara M, Osawa K, Yoshida K, Miyaishi A, Osawa Y et al. (2008) Association of MUTYH Gln324His and APEX1 Asp148Glu with colorectal cancer and smoking in a Japanese population. J Exp Clin Cancer Res 27: 49.
  31. 31. Mitra AK, Singh N, Singh A, Garg VK, Agarwal A et al. (2008) Association of polymorphisms in base excision repair genes with the risk of breast cancer: a case-control study in North Indian women. Oncol Res 17: 127-135.
  32. 32. Pardini B, Naccarati A, Novotny J, Smerhovsky Z, Vodickova L et al. (2008) DNA repair genetic polymorphisms and risk of colorectal cancer in the Czech Republic. Mutat Res 638: 146-153.
  33. 33. Shekari M, Sobti RC, Tamandani DM, Malekzadeh K, Kaur P et al. (2008) Association of genetic polymorphism of the DNA base excision repair gene (APE-1 Asp/148 Glu) and HPV type (16/18) with the risk of cervix cancer in north Indian population. Cancer Biomark 4: 63-71.
  34. 34. Smith TR, Levine EA, Freimanis RI, Akman SA, Allen GO et al. (2008) Polygenic model of DNA repair genetic polymorphisms in human breast cancer risk. Carcinogenesis 29: 2132-2138.
  35. 35. Tse D, Zhai R, Zhou W, Heist RS, Asomaning K et al. (2008) Polymorphisms of the NER pathway genes, ERCC1 and XPD are associated with esophageal adenocarcinoma risk. Cancer Causes and Control 19: 1077-1083.
  36. 36. Zhu R, Wu Y, Lu FJ, Wang AH, Tang JY et al. (2008) Polymorphisms and haplotypes of XRCC1 and APE1 and risk of childhood leukaemia in China: A case-control analysis. Eur J Oncol 13: 187-192.
  37. 37. Agachan B, Kucukhuseyin O, Aksoy P, Turna A, Yaylim I et al. (2009) Apurinic/apyrimidinic endonuclease (APE1) gene polymorphisms and lung cancer risk in relation to tobacco smoking. Anticancer Res 29: 2417-2420.
  38. 38. Gangwar R, Ahirwar D, Mandhani A, Mittal RD (2009) Influence of XPD and APE1 DNA Repair Gene Polymorphism on Bladder Cancer Susceptibility in North India. Urology 73: 675-680.
  39. 39. Ji L (2009) Single-nucleotide polymorphisms in DNA repair gene hOGG1 and APE1 and susceptibility to hepatocellular carcinoma [Master]. Guangxi Medical University.
  40. 40. Liu Y, Scheurer ME, El-Zein R, Cao Y, Do KA et al. (2009) Association and Interactions between DNA Repair Gene Polymorphisms and Adult Glioma. Cancer Epidemiol Biomarkers Prev 18: 204-214.
  41. 41. Lo YL, Jou YS, Hsiao CF, Chang GC, Tsai YH et al. (2009) A Polymorphism in the APE1 Gene Promoter is Associated with Lung Cancer. Risk Cancer Epidemiol Biomarkers Prev 18: 223-229.
  42. 42. Lu J, Zhang S, Chen D, Wang H, Wu W et al. (2009) Functional characterization of a promoter polymorphism in APE1/Ref-1 that contributes to reduced lung cancer susceptibility. FASEB J 23: 3459-3469.
  43. 43. Narter KF, Ergen A, Agachan B, Gormus U, Timirci O et al. (2009) Bladder cancer and polymorphisms of DNA repair genes (XRCC1, XRCC3, XPD, XPG, APE1, hOGG1). Anticancer Res 29: 1389-1393.
  44. 44. Agalliu I, Kwon EM, Salinas CA, Koopmeiners JS, Ostrander EA et al. (2010) Genetic variation in DNA repair genes and prostate cancer risk: results from a population-based study. Cancer Causes Control 21: 289-300.
  45. 45. Brevik A, Joshi AD, Corral R, Onland-Moret NC, Siegmund KD et al. (2010) Polymorphisms in Base Excision Repair Genes as Colorectal Cancer Risk Factors and Modifiers of the Effect of Diets High in Red Meat. Cancer Epidemiol Biomarkers Prev 19: 3167-3173.
  46. 46. Canbay E, Agachan B, Gulluoglu M, Isbir T, Balik E et al. (2010) Possible associations of APE1 polymorphism with susceptibility and HOGG1 polymorphism with prognosis in gastric cancer. Anticancer Res 30: 1359-1364.
  47. 47. Jelonek K, Gdowicz-Klosok A, Pietrowska M, Borkowska M, Korfanty J et al. (2010) Association between single-nucleotide polymorphisms of selected genes involved in the response to DNA damage and risk of colon, head and neck, and breast cancers in a Polish population. J Appl Genet 51: 343-352.
  48. 48. Lv J (2010) Association study and function analysis of APE1/rEF-1 gene promoter polymorphisms with lung cancer and glioma [Doctor]. Fudan University.
  49. 49. Osawa K, Miyaishi A, Uchino K, Osawa Y, Inoue N et al. (2010) APEX1 Asp148Glu gene polymorphism is a risk factor for lung cancer in relation to smoking in Japanese. Asian Pac J Cancer Prev 11: 1181-1186.
  50. 50. Palli D, Polidoro S, D’Errico M, Saieva C, Guarrera S et al. (2010) Polymorphic DNA repair and metabolic genes: a multigenic study on gastric cancer. Mutagenesis 25: 569-575.
  51. 51. Wang M (2010) Genetic variants of APE1 gene contribute to susceptibility to cervical cancer and their mechanisms [Master]. Nanjing Medical University.
  52. 52. Ye C, Huang Z, Zhou C (2010) APE1 D148E, PARP1 V762A and XRCC1 R399Q polymorphisms and genetic susceptibility to colorectal cancer. Wrold. J Gastroenterol: 1275-1279.
  53. 53. Cao Q, Qin C, Meng X, Ju X, Ding Q et al. (2011) Genetic polymorphisms in APE1 are associated with renal cell carcinoma risk in a Chinese population. Mol Carcinog 50: 863-870.
  54. 54. Gu D, Wang M, Wang S, Zhang Z, Chen J (2011) The DNA repair gene APE1 T1349G polymorphism and risk of gastric cancer in a Chinese population. PLOS ONE 6: e28971.
  55. 55. Huang L, Li Q, Xue Y, Li J, He L, et al. (2011) APEl/Ref-1-656A>C and1349T>G polymorphism and risk of children acute lymphoblastic leukemia. Chin J Epidemiol 32: 1179-1180.
  56. 56. Kuasne H, Rodrigues IS, Losi-Guembarovski R, Reis MB, Fuganti PE et al. (2011) Base excision repair genes XRCC1 and APEX1 and the risk for prostate cancer. Mol Biol Rep 38: 1585-1591.
  57. 57. Li Z (2011) Association between DNA repair gene polymorphisms and environmental factors and Hp-associated gastric cancer and duodenal ulcer [Master]. Nanchang University.
  58. 58. Li Z, Guan W, Li MX, Zhong ZY, Qian CY et al. (2011) Genetic polymorphism of DNA base-excision repair genes (APE1, OGG1 and XRCC1) and their correlation with risk of lung cancer in a Chinese population. Arch Med Res 42: 226-234.
  59. 59. Cincin ZB, Iyibozkurt AC, Kuran SB, Cakmakoglu B (2012) DNA repair gene variants in endometrial carcinoma. Med Oncol 29: 2949-2954.
  60. 60. Mandal RK, Gangwar R, Kapoor R, Mittal RD (2012) Polymorphisms in base-excision & nucleotide-excision repair genes & prostate cancer risk in north Indian population. Indian J Med Res 135: 64-71.
  61. 61. Mittal RD, Mandal RK, Gangwar R (2012) Base excision repair pathway genes polymorphism in prostate and bladder cancer risk in North Indian population. Mech Ageing Dev 133: 127-132.
  62. 62. Nakao M, Hosono S, Ito H, Watanabe M, Mizuno N et al. (2012) Selected polymorphisms of base excision repair genes and pancreatic cancer risk in Japanese. J Epidemiol 22: 477-483.
  63. 63. Li Y, Li S, Wu Z, Hu F, Zhu L et al. (2013) Polymorphisms in genes of APE1, PARP1, and XRCC1: risk and prognosis of colorectal cancer in a Northeast Chinese population. Med Oncol 30: 505.
  64. 64. Zhou B, Shan H, Su Y, Xia K, Shao X et al. (2011) The association of APE1 -656T > G and 1349 T > G polymorphisms and cancer risk: a meta-analysis based on 37 case-control studies. BMC Cancer 11: 521.
  65. 65. Cardon LR, Bell JI (2001) Association study designs for complex diseases. Nat Rev Genet 2: 91-99.
  66. 66. Hannah RR, Alexander JS, Michael B (2005) Publication Bias in Meta-Analysis: Prevention, Assessment and Adjustments. Publisher: Wiley: P105.