A Meta-Analysis for Association of XRCC1, XRCC2 and XRCC3 Polymorphisms with Susceptibility to Thyroid Cancer

Background: We conducted a comprehensive meta-analysis to explore the association of polymorphisms at XRCC1, XRCC2 and XRCC3 genes with susceptibility to thyroid cancer (TC). Methods: We searched PubMed, EMBASE, Web of Science, and CNKI for relevant available studies. The pooled odds ratios (ORs) with 95% confidence intervals (CIs) were used to evaluate the strength of the associations. Results: A total of 67 studies including 17 studies with 6,806 cases and 5,229 controls on XRCC1 Arg399Gln, 13 studies with 3,234 cases and 4,807 controls on XRCC1 Arg280His, 13 studies with 2,956 cases and 3,860 controls on XRCC1 Arg194Trp, five studies with 1,287 cases and 1,422 controls on XRCC2 Arg188His, 13 studies with 2,488 cases and 3,586 controls on XRCC3 Thr241Met, and six studies with 1,828 cases and 2,060 controls on XRCC3 IVS5-14 polymorphism were selected. Polled data revealed that the XRCC1 Arg399Gln, Arg280His, Arg194Trp, XRCC2 Arg188His and XRCC3 Thr241Met and IVS5-14 polymorphisms were not significantly associated with an increased risk of TC. Stratified analyses by ethnicity showed that the XRCC1 Arg399Gln polymorphism was associated with TC risk in Caucasians, but not in Asians. Conclusions: Our meta-analysis indicated that the XRCC1 Arg399Gln, Arg280His, Arg194Trp, XRCC2 Arg188His, XRCC3 Thr241Met and IVS5-14 polymorphisms were not associated with risk of TC in the global population. Further well-designed investigations with large sample sizes are required to confirm our results.


Introduction
Thyroid carcinoma (TC) is the most common endocrine malignancy and accounts for 0.5-1.5% of all cancer cases in the United States (Ortega et al., 2004;Joseph et al., 2018). The incidence of TC has increased globally in recent decades, especially among younger adults. Its incidence and mortality rates are varied by 8-12 fold and 2-6-fold, respectively (La Vecchia et al., 2015;Sierra et al, 2016;Sanabria et al., 2018). In the United States of America (USA), deaths from TC alone account for more deaths than all of the other endocrine malignancies with annual incidence of 6.6% between 2000 and 2009 is the highest among all cancers (Davies et al., 2015;Kim et al., 2020;Yan et al., 2020). TC is more common in women than in men, but men are twice as likely as women to die from this cancer (Rahbari et al., 2010). The Papillary Thyroid Carcinoma (PTC) is the most frequent subtype of thyroid malignancy which constitutes approximately >85% of all cases (Schlumberger and Torlantano, 2000;Baloch and LiVolsi, 2018;Joseph et al., 2018). The etiology and development of TC is a result of complex interactions between genetic and environmental factors (Makazlieva et al., 2016;Boi et al., 2017;Nettore et al., 2018). Continuously exposed to a wide range radiation is a well-established risk factor for TC, which such radiation include certain radiation therapy and radiation fallout from power plant accidents of atomic bombs (Yamashita and Suzuki, 2013;Iglesias et al., 2017;Fiore et al., 2019).
There is increasing evidence suggests that damage to human DNA might initiate the cancer, which caused by external agents such as chemical agents, ionizing radiation and UV (Lange et al., 2011;Barnes et al., 2018). To date, several genetic variations that have a fundamental role in the carcinogenesis of different subtypes of TC have been reported (Xing, 2013;Penna et al., 2016). DNA repair is essential for the maintenance of genomic integrity, which is of primary importance in the general and specialized functions of cells, as well as in the prevention of carcinogenesis . Some genes of the X-ray repair cross-complementing (XRCC) family are an essential part of the BER and homologous recombination (HR) DNA repair pathways responsible for DNA double strand breaks caused by normal metabolic processes and/or exposure to ionizing radiation, and have been reported to be associated with development of TC (Namazi et al., 2015;Cannan and Pederson, 2016;Yan et al., 2016). Previous studies have reported that X polymorphisms of XRCC1, XRCC2, XRCC3 DNA repair genes are associated with an increased risk of TC in in different populations (Hu et al., 2013;Jafari Nedooshan et al., 2017). However, no conclusive result has been reported due to the conflicting results among different studies. Therefore, a meta-analysis of all available studies will help to obtain a more convincing result, because some of these studies were based on small sample size, thus, subgroup analysis ethnicity may also yield more meaningful results (Dijkman et al., 2009;Ganeshkumar and Gopalakrishnan, 2013). Here, we performed a meta-analysis of all eligible case-control studies published to date, to assess the association of XRCC1, XRCC2 and XRCC3 polymorphisms with susceptibility of TC globally.

Publication Search
Ethical approval or patient consent was not required because this is a meta-analysis in which all data were extracted from published data. A comprehensive computer search was carried out independently by two authors, in PubMed, Web of Knowledge, Web of Science, Embase, Scientific Information Database (SID), WanFang, VIP, Chinese Biomedical Database (CBD), Scientific Electronic Library Online (SciELO) and China National Knowledge Infrastructure (CNKI) database to collect the case-control studies that investigated the association XRCC1, XRCC2 and XRCC3 genes polymorphisms with TC risk up to October 01, 2020. Combinations of the following keywords were used in the search: (''Thyroid Cancer'' OR ''Thyroid Carcinoma'' OR ''Papillary Thyroid Cancer '' OR ''Follicular Thyroid Cancer'' OR ''Hurthle Cell Cancer'' OR ''Medullary Thyroid Cancer'' OR Anaplastic Thyroid Cancer'') AND (''X-Ray Repair Cross-Complementing Protein I'' OR ''XRCC1'' OR ''rs1799782'' OR ''Arg194Trp'' OR ''rs25487'' OR ''Arg399Gln'' OR ''rs25489 OR ''Arg280His'') AND (''X-Ray Cross Complementing Group II'' OR ''XRCC2'' OR ''Arg188His'' OR ''rs3218536'') AND (''X-Ray Cross Complementing Group III'' OR ''XRCC3'' OR ''Thr241Met'' OR ''rs861539'') AND (''Gene'' OR ''Genotype'' OR ''Allele'' OR ''Polymorphism'' OR ''Single Nucleotide Polymorphisms'' OR ''SNPs'' OR ''Variation'' OR ''Mutation''). In addition, to prevent the loss of any important data, we reviewed the bibliographical references list of retrieved studies, reviews and previous meta-analyses. The whole search process was carried out in English, Chinese, Portuguese, Russian and Persian. When overlapping data on the same cases were included in more than one publication, only the one with the larger sample size was selected.

Inclusion and Excluding Criteria
The studies included in the meta-analysis were required to meet the following criteria: 1) Case-control study of TC cases and healthy subjects; 2) studies evaluated the association of polymorphisms at XRCC1, XRCC2 and XRCC3 genes with TC; 3) provide both genotype and allele distributions inpatients and controls for estimation of combined odds ratios (OR) and 95% confidence intervals (CI); and 4) full text studies on human. Accordingly, Studies were excluded if they: 1) abstracts, reviews, editorials, comments or animal studies; 2) case only studies; 3) linkage studies and family based studies; 4) did not provide the numbers of genotypes; 5) animal and in vitro studies; and 6) contained overlapping data. If the full text article or a study did not published detailed data regarding the genotype distribution in cases and controls, the corresponding authors of the study were contacted for unpublished data.

Data Extraction
Two authors worked independently to extract all data from all eligible studies based on the inclusion criteria. Any disagreement was resolved by further discussion until a consensus about valid data was reached. The publication details collected included: first author's name, year of publication, ethnicity (Asian, Caucasian, African and mixed populations), country of origin, genotyping methods, numbers of cases and controls, frequencies of genotypes in cases and controls, minor allele frequency (MAF) in controls, and Hardy-Weinberg equilibrium (HWE) in controls. The ''mixed'' group means mixed or unknown populations. Moreover, when publications included sample of more than one ethnicity or population, the data was extracted separately according to ethnicities. The publications did not reported necessary data, as well as genotype frequencies; we contacted the corresponding authors by email to request the missing data.

Statistical Analysis
The strength of the association between different polymorphism of XRCC1, XRCC2 and XRCC3 genes and TC risk was estimated by calculating pooled odds ratios (OR) with 95% confidence intervals (CI). The significance of the summary of pooled data was tested using a Z-test in which P-values less than 0.05 were considered to be statistically significant. The association of the XRCC1, XRCC2 and XRCC3 polymorphisms with TC risk was evaluated under models, i.e., allele (B vs. A), homozygote (BB vs. AA), heterozygote (BA vs. AA), dominant (BB+BA vs. AA) and recessive model (BB vs. BA+AA), respectively. The between studies heterogeneity was performed using the chi-square-based Cochrane Q-test, in which P-value less than 0.10 was considered studies on XRCC1, XRCC2 and XRCC3 polymorphisms and TC risk was assessed visually inspecting the Begg's funnel plot for asymmetry and the Egger' linear regression test statistically. Egger`s linear regression test was used to evaluate the symmetry of the funnel plot in order to minimize the subjective influence of the visual inspection assessment, in which bias was considered with P<0.05 in Egger's test. Statistical analyses were performed using Comprehensive Meta-Analysis (CMA) software version 2.0 (Biostat, USA). Two-sided P-values < 0.05 were considered statistically significant.

Characteristics of included studies
The selection process of eligible studies is presented in Figure 1. A total of 515 potentially relevant articles were preliminarily identified though a systematic publication significant. In addition, we have used I2 to statistically measure the heterogeneity and indicate the percentage of variance of the heterogeneity. A fixed-effect model (Mantel-Haenszel method) was used to pool ORs and 95% CI when there was no significant heterogeneity. Otherwise, a random effects model (the DerSimonian and Laird method) was used. The Pearson's χ 2 test was applied to test the Hardy-Weinberg equilibrium (HWE) in healthy controls with the significance set at P<0.05. Sensitivity analysis was performed by iteratively omitting one study at a time to determine the effects of individual study on overall data and stability of the results. Moreover, sensitivity analysis was performed by removing those studies did not in agreement with HWE in control groups.   DOI:10.31557/APJCP.2021.22.7.2221 XRCC and Thyroid Cancer   (Zhu et al., 2004(Zhu et al., , 2018Sturgis et al., 2005;HX et al., 2006;Siraj et al., 2008;Chiang et al., 2008;Bastos et al., 2009;Ho et al., 2009;Akulevich et al., 2009;Fard-Esfahani et al., 2011;García-Quispes et al., 2011;Ryu et al., 2011;Santos et al., 2012;Fayaz et al., 2013;Wang et al., 2015;Halkova et al., 2016;Yuan et al., 2016;Yan et al., 2016;Sarwar et al., 2017;Adampourezare et al., 2017;Bashir et al., 2018). Detailed characteristics and genotype distribution of eligible studies are summarized in Table 1 (Table 1). Table 3 presents the main results of the meta-analysis of the XRCC1 Arg399Gln, Arg280His and Arg194Trp polymorphisms and TC risk. Pooled data revealed that the XRCC1 Arg399Gln, Arg280His and Arg194Trp polymorphisms were not significantly associated with an increased risk of TC in the global population ( Figure  2). When stratified by ethnicity, the XRCC1 Arg399Gln polymorphism was associated with risk of TC in Caucasians under two genetic models, i.e., allele (A vs. G: OR=0.334, 95% CI 0.789-0.980, p=0.020) and dominant (AA vs. GG: OR=0.869, 95% CI 0.760-0.993, p=0.040), but not in Asians. Subgroup analyses by ethnicity still did not find a significant for association of XRCC1 Arg280His and Arg194Trp polymorphisms and TC risk (Table 3). Table 2 listed the main results of the meta-analysis of XRCC2 Arg188His polymorphism and TC risk. We pooled all the five case-control studies to evaluate the association of XRCC2 Arg188His polymorphism with TC risk. The pooled results showed that XRCC2 Arg188His polymorphism did not significantly associate with TC risk under all five genetic models (Figure 3). When, subgroup analyses performed according to ethnicity still did not find significant association between XRCC2 Arg188His polymorphism and TC risk in Asians and Caucasians (Table 3).

XRCC3 Polymorphisms
The summary for the association of the XRCC3 Thr241Met and IVS5-14 polymorphisms with TC risk are summarized in Table 3. Pooled data revealed that the XRCC3 Thr241Met and IVS5-14 polymorphisms were not significantly associated with risk of TC under all five genetic models ( Figure 4). When, subgroup analyses performed according to ethnicity still did not find significant association between XRCC3 Thr241Met polymorphism and TC risk in Asians and Caucasians (Table 3).

Test of Heterogeneity
Significant heterogeneity existed in all of the genetic models for XRCC1 Arg399Gln, Arg280His, Arg194Trp, XRCC3 Thr241Met and IVS5-14 polymorphisms (Table  3). Thus, we performed subgroup analyses by ethnicity to find the possible source of heterogeneity. Results showed that Caucasians descent subjects have not overall effect on the heterogeneity, but the selected Asian descents were extremely heterogeneous.

Sensitivity Analysis
Sensitivity analyses were performed after sequentially removing each eligible study to assess the stability of our results. This test is regarded as an indispensable step for analyzing multiple criteria. The results showed that the significance of the pooled ORs was not influenced by any single study under all five genetic models for XRCC1, XRCC2 and XRCC3 polymorphisms, indicating that our results were highly stable. Moreover, we performed sensitivity analysis by excluding those studies did not in agreement HWE for XRCC1 Arg399Gln, Arg280His, Arg194Trp, XRCC3 Thr241Met and XRCC3 IVS5-14 polymorphisms. Similarly, after excluding those studies the results indicated no significant alteration in the pooled ORs.

Publication Bias
We used the Visual inspection of funnel plot and the Egger's weighted regression tests to assess the publication bias of eligible literatures for XRCC1, XRCC2 and XRCC3 polymorphisms and TC risk. Visual inspection of the funnel plots did not show any evidence of publication bias for XRCC1 Arg399Gln, Arg280His, Arg194Trp, XRCC2 Arg188His, XRCC3 Thr241Met and IVS5-14 polymorphisms ( Figure 5). Moreover, the Egger test, which was used to provide statistical evidence of funnel plot symmetry, did not show any significant publication bias in this meta-analysis (Table 3).

Discussion
Human XRCC1 gene is mapped to chromosome    (Li et al., 2013). The XRCC1 protein has no known catalytic activity, but serves an important component of the base excision repair (BER) pathway via its role as a central scaffolding protein physically associated with DNA ligase III at its COOH terminus (Li et al., 2012). More than 300 validated polymorphisms in the human XRCC1 gene are listed in the dbSNP database, of which, the most extensively studied SNPs are Arg399Gln (exon 10), Arg280His (exon 9) and Arg194Trp (exon 6) polymorphisms in different cancer (Li et al., 2012, Li et al., 2013Qi et al., 2014). Our results revealed that the XRCC1 Arg399Gln, Arg280His and Arg194Trp polymorphisms were not significantly associated with risk of TC in the global population. However, subgroup analysis showed that the XRCC1 Arg399Gln polymorphism was associated with risk of TC in Caucasians, but not in Asians. To date, several meta-analyses have been performed to undertake the association of polymorphisms in XRCC1 in development of TC. Human XRCC2 gene is paralogue of RAD51 plays a pivotal role in the homologous recombination repair (HRR) machinery, maintenance of the genome integrity and the control of genomic rearrangement processes causes to the chromatid breaks (Kamali et al., 2017). XRCC2 gene is located on human chromosome 7q36.1, consists of three exons, which are distributed 29 DNA repair over a 30 kb region. In exon 3, an Arg188His polymorphism (rs3218536) has been identified on the coding region of XRCC2 as potential cancer susceptibility loci in recent studies, although association results are controversial. However, the potential phenotypic effects  Human XRCC3, also known as CMM6, is a member of the RecA/Rad51-related protein family that participates in HRR to maintain chromosome stability which was originally identified by its ability to complement the DNA repair defect (Duarte et al., 2005;Sobhan et al., 2017). Human XRCC3 gene is located on chromosome 14q32.3, contains 10 exons (its seven exons lie in the region taking 13.5 kbp) and spans 21 kbp length (Ali et al, 2016;Liu et al, 2019). In this meta-analysis, our pooled data showed that the XRCC3 IVS5-14 and Thr241Met polymorphisms were significantly associated with an increased risk of TC in the overall population. Moreover, subgroup analysis showed that there was no a significant association between the XRCC3 IVS5-14 and Thr241Met polymorphisms and an increased risk of TC. Unlike our results, Lu et al., in a meta-analysis of eight studies with 963 TC cases and 1,942 controls reported that the XRCC3 Thr241Met polymorphism was associated with the risk of TC in the global population, but they did not observe significant association in by ethnicity (Lu et al., 2015). On the basis of availability of five more studies with 2,589 cases and 3,596 controls on XRCC3 Thr241Met polymorphism and TC, our results more reliable and powerful results than the previous meta-analysis.
The present meta-analysis has some novelty and advantages. First, to the best of our knowledge, this study was the first meta-analysis to comprehensively evaluate the association of XRCC2 Arg188His polymorphism with susceptibility to TC. Second, our results were inconsistent with the previous meta-analysis on XRCC3 Thr241Met polymorphism association with TC risk might be due to including large sample size. Finally, no publication bias was found in the present study and sensitivity analysis also indicated that no single study yield obvious impact on the pooled results, which indicating that the results of the present meta-analysis are statically robust.
Despite above mentioned advantages, the current meta-analysis has some limitations which should be addressed. First, the sample size is still relatively small, which might not enough statistical power to explore the real association of the XRCC2 Arg188His and XRCC3 IVS5-14 polymorphisms with TC risk, which leads to the improper publication bias for these polymorphisms. Second, in the meta-analysis all of the included studies were on the Caucasian and Asians, and there was no study in African and mixed populations among the eligible studies. Therefore, need to further studies on a large scale on African and mixed populations to verify this result. Third, the study might have experienced the publication bias due to the inclusion of English and Chinese literature, which could have limited the published evidences. Fourth, the control group of several studies was not in accordance with HWE, which may be attributed to the reason as genotyping error. However, deletion of those studies did not change the results of quantitative synthesis, suggesting the robustness of results. Fifth, our pooled ORs were based on un-adjusted data for potential covariates such as age, sex, lifestyle, exposure and environmental factor, which might have affected the accuracy of the results, though no sufficient information available for most of studies included in the meta-analysis. Finally, TC is a multi-factorial disease from complex interactions between environmental factors and genetic factors. In this metaanalysis, we had insufficient data to conduct an evaluation of such interactions for the role of XRCC1, XRCC2 and XRCC3 polymorphisms and factors in TC development.
In summary, the present meta-analysis suggested that the XRCC1 Arg399Gln, Arg280His, Arg194Trp, XRCC2 Arg188His, XRCC3 Thr241Met and IVS5-14 were not significantly associated with an increased risk of TC in global population. However, subgroup analyses by ethnicity showed that the XRCC1 Arg399Gln polymorphism was associated with risk of TC in Caucasians, but not in Asians. Taking into account the aforementioned limitations, further studies are highly needed in the future.

Author Contribution Statement
Conceived and designed the study and experiments: MM, SAD, SMT. Performed the experiments: FA and JJN Analyzed the data: HN and SAD. Contributed reagents/ materials/analysis tools: SHS and SK. Wrote the paper: HN, MM and FA. All authors reviewed the manuscript.