Association between the XRCC1 Arg194Trp Polymorphism and Glioma Risk: an Updated Meta-analysis

Gliomas are the most common type of primary brain tumors (Ricard et al., 2012), with an incidence rate of approximately 6/100,000 per year worldwide. Despite the advances in neurosurgery and chemotherapy, median survival of only 12 to 15 months among patients in the United States with glioblastoma, the most common type of glioma (Wen et al., 2008). Nowadays, the cause of glioma is still unknown and the etiology has been poorly understood, and may be multifactorial resulting from the interaction of intrinsic and environmental factors (Connelly et al., 2007, Bondy et al., 2008). The only established environmental risk factor is exposure to therapeutic or high-dose ionizing radiation (Schwartzbaum et al., 2006; Bondy et al., 2008; Ostrom et al., 2011). X-ray repair cross complementing group 1 (XRCC1) acts as a scaffolding protein that functions in the repair of base excision and DNA single-strand breaks, the two most common repair pathways in cellular DNA (Caldecott et al., 1995). XRCC1 interacts with a number of proteins crucial to the BER/SSBR pathways, including OGG1, NEIL2, NTH1, MPG, UNG2, AP endonuclease-1 (APE-1), poly (ADPribose) polymerase, DNA polymerase β, and DNA ligase 3 (Caldecott et al., 1995; Dianov et al., 1999; Thompson et al., 2000; Vidal et al., 2001; Marsin et al., 2003; Campalans

Recently, Jiang et al. (2013) reported that XRCC1 Arg194Trp polymorphism might have no influence on the susceptibility of glioma; However, only four literatures were included in this meta-analysis. Subsequently seven molecular epidemiologic studies on the association between this polymorphism and glioma risk also presented contradictory results. Here, we update previous metaanalyses, with additional data to assess the effect of XRCC1 Arg194Trp polymorphism on glioma incidence. In this meta-analysis, we aimed to obtain outline risk evaluates for the XRCC1 Arg194Trp associated with glioma risk.

Materials and Methods
We searched the electronic databases Web of Science, PubMed and EMBASE using such terms ("glioma" or "gliomas" or "brain cancer"), ("XRCC1" or "X-ray repair cross-complementation group 1" or "DNA repair gene" and "polymorphism or variant or variation" ) (last search was updated on February12, 2014).
The inclusion criteria of this meta-analysis were: 1) XRCC1 Arg194Trp polymorphism and glioma; 2) sufficient maternal genotype data for estimating an odds ratio (OR) with a 95% confidence interval (CI); and 3) published in English. The criteria for the exclusion of studies are as follows: 1) not relate to the XRCC1 Arg194Trp polymorphism and glioma;2) not a primary case-control study; 3) no usable or sufficient maternal genotype data reported.

Data collection
The first author, publication year, country of origin, ethnicity, sources of controls, genotyping method, frequency of Trp-allele in controls, number of genotyped cases and controls were collected independently by two authors (XC and CP) in Table 1.

Statistical methods
First, we assessed HWE for the controls in each study. X 2 test of heterogeneity was calculated in comparison with pooled articles, when p value was >0.10 we used fixed-effects model (Mantel-Haenszel); In contrast, the random-effects models (DerSimonian and Laird) was used. Subgroup analyses were also conducted by ethnicity, study design, genotyping method and HWE. Also, the sensitivity analyses and publication bias was performed. In brief each time a single article was removed, then we analysed remain articles respectively. The methods of Egger et al. and Begg et al. were to test the publication bias. The result consists of the Begg's funnel plot and Egger's test. All statistical analyses were performed using the STATA software version 11 (Stata Corporation, College Station, TX). Two-sided P values less than 0.05 were considered statistically significant

Literatures
The characteristics of the selected studies are listed in Table1.There twenty-one studies were meet our search terms, and 12 eligible studies were finally included in. Totally, 5407 patients and 7715 healthy persons were for meta-analysis. Among our analysis seven studies of Study population were Asians and four were Caucasians, three studies were population-based controls and eight studies were hospital-based controls. The distributions of the genotypes in the control groups in 7 studies were not in HWE. All of included articles were able to analyse for the allel model, additive model, dominant model and recessive model. The major baseline characteristics of the 12 eligible publications were reported in Table 1.

Subgroup analysis
The similar association was discovered in the subgroup analyses. In the subgroup analyses were based on Ethnicity and sources of control. The results were robust, which did not vary materially after we excluded the study with controls not in HWE. Following significant results were to describe.
Among subgroup of ethnicity, only in Asian existed significant results were in following genetic models : additive model ( Arg194Trp polymorphism was no association with glioma. As well in populaiton-based on controls, similar significant results were in population-based on controls from hospital. The detailed information was in Table 2. Additionally, when the Genotyping method was MassARRAY, all statistic models presented significantly increased risks.

Test for heterogeneity, sensitivity analyses and publication bias
Pooled comparisons and subgroup analyses were examined the heterogeneity. In allel and dominant models, among pooled analysis the heterogeneity of P values were all <0.1, the results were shown in Table 2. Therefore, we performed the source of heterogeneity among Ethnicity, sources of control, genotyping method and HWE. When we performed the sensitivity analyses, no matter overall analyses and the subgroup analyses ORs was not altered, suggesting that our results were stability and liability statistically. Also the sensitivity result was in Figure 3

Discussion
Glioma is generally considered to be a geneenviroment interaction disease, and a better understanding of the mechanism of glioma will help us find better ways to prevent, diagnose, or treat glioma. At present, notwithstanding some risk factors have been found, the etiology of glioma is still poorly understood (Kishida et al., 2012, Marumoto et al., 2012. However, as we all know that genetic factors play crucial roles in the occurrence of glioma (Melin, 2011;von Deimling et al., 2011).  DOI:http://dx.doi.org/10.7314/APJCP.2014.15.17.7419 Association between the XRCC1 Arg194Trp Polymorphism and Glioma Risk: an Updated Meta-analysis to make early diagnosis, predict patients outcome, or carry out individualized or personalized therapy. Unfortunately, up to now few genetic biomarkers has been identified as good biomarkers for glioma patients. To find some glioma relevant genetic biomarkers is of the most importance to improve the prognosis.

Confirmed of biomarkers of genetic factors could expect
As we all know, DNA repair genes could maintain the genome integrity, and thus DNA repair genes polymorphisms are potential candidates which can modify the development of gliomas. XRCC1 is one of the most important DNA repair genes responsible for BER pathway and DBS caused by IR. The XRCC1 Arg194Trp polymorphism is located in an evolutionary conserved linker region, makes the chances of occurrence of chromosomal breaks highly increased (Monaco et al., 2007). Up to now, a number of articles have been performed to address the association between XRCC1 Arg194Trp polymorphism and the risk of gliomas, but yielded conflicting results. Because of the above-mentioned contradictory results from relatively small articles underpowered to detect the effects and previous meta-analysis indicated that XRCC1 Arg194Trp polymorphism was no association with the development of glioma an updated meta-analysis should be a proper way to obtain a more definitive conclusion.
Although previous meta-analyses have confirmed that the results did not showed any association between XRCC1 Arg194Trp polymorphism and glioma risk for all genetic models (Jiang et al., 2013), even in subgroup analyses based on the source of controls, ethnicity and histological subtype. After that, published data regarding the association between Arg194Trp polymorphism and glioma risk were inconsistent. To derive a more accurate estimation of the association between XRCC1 Arg194Trp polymorphism and glioma risk, we performed a meta-analysis. This meta-analysis, including a total of 5407 glioma cases and 7715 controls from 12 casecontrol studies, examined the potential association of the polymorphisms of the DNA repair gene XRCC1 Arg194Trp with glioma risk. We observed a fixed overall 25.9% increased risk of glioma for the Trp allel of the Arg194Trp polymorphism, compared with the wild allel (OR, 1.259; 95%CI, 1.045-1.517). Similarly, for the Arg 194Trp, the variant genotypes (TrpTrp), compared with the wild-type homozygote (Arg/Arg), were associated with a significantly increased glioma risk (OR, 2.108; 95%CI, 1.593-2.789) for ethnicity types without between-study heterogeneity. However, due to the presence of marginal statistical evidence and small sample size for Arg194Trp, our result as regards this polymorphism should always be regard as preliminary. However, our analysis shows that even if a general variant in the functional region of a conclusive meaningful gene had an effect on human disease, such as glioma, it may play only a tiny role in the development of glioma, which is conform to the characteristics of low-penetrance genes (Lohmueller et al., 2003).
Furthermore, three recent meta-analyses by Luo et al. (2013), Pan et al. (2013) and Xu et al. (2013) were evaluated the association between Arg194Trp polymorphism and glioma risk, which was basically in consist with our meta-analysis results that Arg194Trp polymorphism may contribute to the susceptibility of glioma, particularly in Asians, but not in Caucasians. It is notable that given the specific multiplicity of possible comparisons and the inescapable adaptation of choosing, associations may have been detected by chance alone. Some articles have been proposed for evaluating correlations between genetic polymorphisms and disease (Freely associating, 1999). The claim was that studies "ideally should have large sample sizes, small P values, report associations that make biological sense, and alleles that affect the gene product in a physiologically meaningful way" (Hu et al., 2005). The scientific hypotheses and sample size of the study are crucial to know the ratio of false-positive findings of meta-analysis that are attributable to constituent studies with selection bias from publication, poor study design, and nondifferential misclassification errors (Wacholder et al., 2002).
One study conducted in region of Europe with 700 glioma patients and 1556 controls reported that no association between the Arg194Trp polymorphism and glioma cancer risk (Kiuru et al., 2008). The other studies in USA consisted of a total sample size (1514 cases and 2755 controls) showed that Arg194Trp did not confer an effect on glioma (Liu et al., 2009;McKean-Cowdin et al., 2009;Rajaraman et al., 2010). Two of articles did not contain sex, age and other match statistic parameters, whereas since 2012, published articles included detailed statistic parameters such as smoking, drinking, cancer history of first relatives and IR exposure, which suggested that support the current meta-analysis that XRCC1 Arg194Trp may play a role in individual susceptibility to glioma.
In addition, between-study heterogeneity is a potential problem which was not avoidable. Despite several differences in the studies about ethnicity, sample sizes, source of controls, and genotyping method, we didn't observe significant heterogeneity between studies for the Arg194Trp polymorphism. Importantly, we carefully performed sensitivity analysis according to sample size and leave-one-out analysis, conducted different conclusions with the previous meta-analysis. In view of this, the results of our meta-analysis, substantially, are sound and reliable.
Similar to other meta-analyses, our study also has a few potential limitations. First, owing to lack of adjusted variables the present meta-analysis was based primarily on unadjusted effect estimates and CIs, thus the effect estimates were relatively imprecise, a more accurate analysis could be conducted if adjusted variables were available in all articles. Second, quite small sample size existed for several subgroup analyses, such as source of controls from population. Third, glioma is known as a multifactor disease, due to lack of detailed data, such as environmental factors, physical inactivity and dietary state factors, thus the gene-gene and gene-environment interactions were not addressed in this meta-analysis. Fourth, several articles indicated that demographic parameters are not well adjusted statistically (Kiuru et al., 2008;McKean-Cowdin et al., 2009). Fifth, misclassifications of genotypes may also impact the results because cases were not verification by other gold standard methods in several studies, and the quality control of genotyping was also not well-verification in some articles. Lastly, although we did not discover publication bias, selection bias may exist because only literatures published in English were included.
In conclusion, our current study support that XRCC1 Arg194Trp polymorphism may contribute to individual susceptibility of glioma. To further evaluate genegene and gene-environment interactions on XRCC1 polymorphisms and glioma risk, thousands of subjects and tissue-specific biochemical characterizations are required.