Associations Between XPD Lys751Gln Polymorphism and Leukemia: A Meta-Analysis

Objectives: The aim of the present study was to define the potential relationship between xeroderma pigmentosum group D (XPD) Lys751Gln polymorphisms and the risk of leukemia. Methods: A comprehensive search of Pubmed, Web of Science, EBSCO, the Cochrane Library and China National Knowledge Infrastructure was conducted to identify original articles published before March 2017 concerning the association between XPD Lys751Gln polymorphisms and leukemia risk. A literature quality assessment was performed using the Newcastle-Ottawa Scale. Heterogeneity across studies was assessed using I2 statistics. Random- or fixed-effects models were used to calculate pooled odds ratios (ORs) in the presence or absence of heterogeneity, respectively. Sensitivity analysis was used to assess the influence of individual studies on the pooled estimate. Publication bias was investigated using funnel plots and Egger’s regression test. All data analyses were performed using Stata 14.0 and Revman 5.3. Results: Fourteen studies with a total of 7525 participants (2,757 patients; 4,768 controls) were included in this meta-analysis. We found that XPD Lys751Gln polymorphisms significantly increased the risk of developing leukemia in both dominant OR = 1.21, 95%CI [1.10–1.35], P ≤ 0.001) and heterozygote (OR = 1.22, 95%CI [1.09–1.36], P ≤ 0.001) model. An allele model showed a borderline significant increase in leukemia risk (OR = 1.13, 95%CI [1.00–1.27], P = 0.05). A subgroup analysis revealed a consistent association between XPD Lys751Gln polymorphisms and leukemia risk for some genetic models in Caucasian populations, adult or chronic groups, and in almost all models of childhood or acute groups. Conclusion: Our results indicate that XPD Lys751Gln polymorphism increases the risk of leukemia, especially in childhood and acute cases.


INTRODUCTION
Leukemia, a common malignant disease of the hematopoietic system (Jiang et al., 2014), can be classified on the basis of speed of disease progression and cell cytogenetics into four common subtypes: acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), chronic myeloid leukemia (CML) and chronic lymphocytic leukemia (CLL) (Arber et al., 2016). The etiology and mechanisms underlying of leukemogenesis are still unclear, although radiation, smoking, obesity and exposure to chemical carcinogens are considered high risk factors (Larsson and Wolk, 2008;Fircanis et al., 2014;Malagoli et al., 2016;Nikkila et al., 2016). Nevertheless, only a small proportion of people exposed to these risks develops leukemia, suggesting that hereditary factors may play a critical role in leukemia carcinogenesis (Li et al., 2016;Huang and Ovcharenko, 2017).
Decreased efficiency of DNA repair is considered a crucial event in carcinogenesis (Hoeijmakers, 2001). Recently, a series of studies revealed that reduced DNA repair, leading to chromosomal aberrations and genomic instability, is a major contributor to the pathogenesis of leukemia (Das-Gupta et al., 2000;Esposito and So, 2014). The xeroderma pigmentosum group D (XPD) gene, also known as ERCC2, encodes a 5 -3 superfamily 2 helicase that plays a key role in unwinding the DNA double helix around damaged DNA during nucleotide excision repair (NER) (Kuper et al., 2012;Constantinescu-Aruxandei et al., 2016). Because of the biological significance of XPD, XPD polymorphisms have been extensively studied in different malignant diseases, such as pancreatic (Wu et al., 2017), colorectal (Ni et al., 2014) and gallbladder (Srivastava et al., 2010) cancers.

Search Strategies
A computerized search of Pubmed, Web of Science, EBSCO, the Cochrane Library and China National Knowledge Infrastructure (CNKI) up to March 2017 was conducted using the following search strategy: ("XPD" or "ERCC2" or "xeroderma pigmentosum group D"), and ("polymorphism" or "variant" or "mutation"), and "Leukemia." The search was restricted to English and Chinese language publications. A manual search of the references in the retrieved articles and relevant reviews was also conducted. A flowchart of information pertaining to identification, screening, eligibility, and the final selected datasets was constructed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines (Moher et al., 2009).

Inclusion and Exclusion Criteria
Studies that investigated the association between XPD Lys751Gln polymorphism and leukemia risk were included. The inclusion criteria were (1) case-control study design; (2) available genotype information of the XPD Lys751Gln polymorphism; (3) evaluation of the XPD gene polymorphism and the risk of leukemia; and (4) the distribution of genotypes among the controls was in agreement with the Hardy-Weinberg equilibrium (HWE). The major criteria for exclusion were: (1) duplication of earlier publications (for studies using the same sample in different publications, only the most complete information was included following careful examination), (2) unpublished papers, dissertations, conference articles and reviews, (3) family-based studies of pedigrees.

Data Extraction
Data from each eligible study were extracted into Excel, and included the country of origin and ethnicity of each study population, age group (adult or childhood), subtypes of leukemia, genotyping method, numbers of cases and controls, numbers of cases and controls in the XPD Lys751Gln genotypes and results of the HWE test.

Study Quality Assessment
The quality of the included studies was assessed by two reviewers in accordance with the Newcastle-Ottawa Scale (NOS) (Stang, 2010), which is used to assess the quality of observational studies. Discrepancies were reported and settled by a third party. Three major aspects of study quality were scored: (1) selection of the study groups (0 ± 4 points); (2) determination of the exposure of interest in the studies (0 ± 3 points); and (3) the quality of the adjustment for confounding variables (0 ± 2 points). A study could be scored with a maximum of one star for each item numbered within the categories of Selection and Exposure, while at most two stars could be allocated to Comparability. A higher score represents a greater quality of the study methodology. A score equal to or higher than 6 was considered to indicate high study quality.

Data Analysis
The combined odds ratios (ORs) and 95% confidence interval (CI) were used to evaluate the strength of the association with the risk of leukemia. Pooled ORs were performed for allelic comparison (a vs. A), dominant (aa + Aa vs. AA), recessive (aa vs. Aa + AA) and codominant (aa vs. AA and Aa vs. AA) models ("a" and "A" represents the mutant allele and the wild-type allele, respectively). Heterogeneity among the included studies was assessed using the I 2 statistic. A random-effects model or fixed-effects model was used to calculate the pooled OR in the presence or absence of heterogeneity, respectively. To detect possible sources of heterogeneity and potential differences among subgroups, meta-regression and subgroup analyses were carried out with the stratification of different ethnicities, age groups and subtypes of leukemia. The significance of the pooled OR was determined through a Z-test, and p < 0.05 was considered to be statistically significant. Publication bias was investigated using funnel plots and Egger's regression test. We also conducted sensitivity analysis to test the robustness of associations by sequentially omitting each of the included studies one at a time. All the data analysis was performed using STATA software 14.0 (StataCorp, College Station, TX, United States) and Review Manager 5.3 (Cochrane Collaboration, Oxford, United Kingdom).

Literature Search and Study Characteristics
We used several search criteria to include or exclude reported studies on the relationship between XPD polymorphism and leukemia (Figure 1). A total of 14 studies (2,757 cases and 4,768 controls) concerning XPD Lys751Gln polymorphism were included in the final evaluation (Table 1). A quality assessment of the individual studies showed that the Newcastle-Ottawa scale score ranged from 6 to 8, indicating that the quality of the methodology was generally good ( Table 2).
HWE P-value in the control group, P > 0.05 indicates that the participants in the control group met the HWE.
Frontiers in Genetics | www.frontiersin.org  risk (Gln vs. Lys: I 2 = 50%, P = 0.05, Figure 6). Moderate heterogeneity (I 2 : 43%-51%) was found in the no-association model group. To explore the source of this heterogeneity, a meta-regression analysis was conducted. The results revealed that the heterogeneity was not associated with ethnicity, age, clinical subtype or detection method (p ≥ 0.05 in all genetic models). We further explored the source of heterogeneity by removing one study each time. The results showed that Matullo et al. (2006) was one of the central sources of the heterogeneity (its inclusion increased heterogeneity by 12%-24%). No publication bias was found in any of the models. Sensitivity analysis suggested that with exception of the allele model, the results were stable and reliable.

DISCUSSION
The NER pathway is a highly conserved DNA repair mechanism that removes bulky intra-strand adducts created by agents such as UV radiation and certain chemicals, including several commonly used chemotherapy agents (Scharer, 2013). Genetic polymorphism in DNA repair genes may cause variation in DNA repair capacity, which in turn can lead to cumulative genotoxic damage and increased susceptibility to cancer (Douzi et al., 2015). As an important component of NER, XPD is an evolutionarily conserved ATP-dependent DNA helicase that plays an essential role in DNA repair (Kim et al., 2016). Several studies have suggested that individuals with polymorphisms in XPD or other NER pathway genes may have an increased risk of cancer (Paszkowska-Szczur et al., 2013;He et al., 2016). Moreover, several molecular epidemiological studies have found an association between XPD polymorphism and leukemia risk in diverse populations. However, the results are inconsistent and even contradictory. We therefore conducted a meta-analysis to globally evaluate the potential relationship between XPD polymorphism and leukemia.
Our results indicate that XPD Lys751Gln polymorphism significantly increases the overall leukemia risk in dominant and heterozygote models, but not in an allele model or homozygote model. The results suggest that heterozygous mutations but not homozygote mutations of XPD (Lys/Gln) may increase genetic susceptibility to leukemia. This may be the result of a higher rate of heterozygous versus homozygous mutations (the ratio between heterozygous and homozygous mutations is 2.27 ∼ 56 in the control and 1.71 ∼ 29 in leukemia). A subgroup analysis by ethnicity showed the same result in a Caucasian population. The exact mechanism underlying the association between the susceptibility to different tumors and XPD Lys751Gln polymorphism is currently unknown. XPD is a 5 -3 superfamily 2 DNA helicase that opens damaged DNA for bulky lesion repair in NER. The interaction of the C-terminal domain of XPD with the p44 helicase activator protein is critical for both helicase activity and the stability of the Transcription Factor II H (TFIIH) complex, which is essential for RNA polymerase II-mediated transcription initiation and the NER (Liu et al., 2008). The XPD C-terminal Lys751Gln polymorphism may alter the structure of the C-terminal domain, there by blocking critical interaction with p44 and creating a destructive TFIIH conformation, which subsequently reduces DNA repair activity (Lunn et al., 2000;Fan et al., 2008;Monaco et al., 2009). Furthermore, by stratifying the data by age and disease subtype, we found that XPD Lys751Gln polymorphism significantly increased leukemia risk in almost all models of childhood and acute disease. The occurrence and development of leukemia appeared to be regulated by genetic and environmental factors. In children with acute leukemia, malignancy manifests with a short latency period, and therefore there is not enough exposure time to allow the initiation of a long carcinogenic process. Unlike children, adults usually develop cancer because of the cumulative effect of environmental exposure during their life (Brisson et al., 2015). Therefore, we speculate that genetic polymorphism is more important for childhood and acute leukemia.
Compared with previous reports Wu et al., 2014), the present study has the following advantages: (1) it analyzed the association between XPD Lys751Gln polymorphism and both acute and chronic leukemia; (2) a total of 14 studies (2,757 cases and 4,768 controls) were included in this analysis, there by increasing the statistical power of the analysis. (3) Strict literature inclusion and exclusion criteria were enforced, the Ozdemir's study was excluded from our analysis because of a lack of discerning information between ALL and Burkitt lymphoma patients. However, our research also has some limitations: (1) the number of original studies included in the meta-analysis is relatively small, especially in the subgroup analysis. Future studies on ethnicity, age and subtype of leukemia are needed to further corroborate our findings. (2) Heterogeneity, which can greatly affect the conclusions of metaanalyses, was high in some models. Our results showed that moderate or higher heterogeneity was found in some models, which showed no association with leukemia risk. Matullo et al. (2006) was a major source of this heterogeneity, probably to because of the significant differences in the number of cases and controls.

CONCLUSION
Our meta-analysis demonstrates that XPD Lys751Gln polymorphism significantly increases overall leukemia risk in dominant and heterozygote models, and that this polymorphism is significantly associated with almost all genetic models of childhood and acute leukemia.

AUTHOR CONTRIBUTIONS
XH and Y-BD designed this study. XL, YC, and JS searched databases and collected the full-text papers. MW and BZ extracted and analyzed the data and wrote the manuscript. EZ, YL, XH, and Y-BD reviewed the manuscript.