Meta-Analysis Results on the Association Between TP53 Codon 72 Polymorphism With the Susceptibility to Oral Cancer

Objectives: TP53 is an important tumor suppressor gene to maintain genomic integrity, and its mutations increase the susceptibility to oral carcinoma. Previous published studies have reported the relation of TP53 codon 72 polymorphism with the risk of oral carcinoma, but the results remain controversial and inconclusive. Methods: We therefore utilized meta-analysis based on a comprehensive search in PubMed, EMBASE, and Google of Scholar databases up to August 19, 2017. Results: Total 3,525 cases and 3,712 controls from 21 case-control studies were selected. We found no significant association between TP53 codon 72 polymorphism and oral carcinoma susceptibility in all genetic contrast models, including subgroup analysis based on control source and ethnicity. Furthermore, TP53 codon 72 polymorphism was not significant associated with oral carcinoma susceptibility in tobacco or alcohol use, and HPV infection status. Our results were confirmed by sensitivity analysis and no publication bias was found. Conclusions: Taken together, our data indicate that TP53 codon 72 polymorphism is not associated with the susceptibility to oral carcinoma.


INTRODUCTION
Based on the GLOBOCAN2012 investigations, oral carcinoma is regarded as one of the most common causes of cancer related morbidity and mortality, contributing to 3.8% of all cancer cases and 3.6% of cancer related deaths (Warnakulasuriya, 2009;Ferlay et al., 2015;Shield et al., 2017). The long-term survival rate of oral carcinoma is <50% despite improved treatment schedules such as surgery, radiation and chemotherapy (Coleman et al., 2008;De Angelis et al., 2014). Oral carcinoma is highly associated with tobacco smoking, alcohol consumption and the exposure to a variety of exogenous or endogenous carcinogens (Petti, 2009). However, the etiology of oral carcinoma remains poorly understood. Furthermore, not all individuals exposing to these risk factors are subject to oral carcinoma, and additional genetic factors may also contribute to oral carcinoma susceptibility (Chen et al., 2010;Anantharaman et al., 2011;Niu et al., 2012).
The human TP53 is well-known tumor suppressor gene and plays an important role in DNA damage response by inducing cell cycle arrest or apoptosis (Slee et al., 2004;Harris and Levine, 2005). TP53 mutation is frequently found in human tumors, including oral carcinoma (Olivier et al., 2002). A common single nucleotide polymorphism at TP53 codon 72 is crucial for its tumor suppressor function (Suzuki and Matsubara, 2011). Several meta-analysis demonstrated that TP53 codon 72 polymorphism was associated with the susceptibility to a variety of cancers, such as colorectal cancer (Du et al., 2017), esophageal cancer (Steccanella et al., 2017), nasopharyngeal cancer (Zhuo et al., 2009a), and non-Hodgkin lymphomas (Xu et al., 2017). Själander et al. demonstrated that the distribution of TP53 genotypes differed among different ethnicities, which is a notable confounding factor in carcinoma risk (Själander et al., 1995). Tobacco and alcohol use are known risk factors for oral carcinoma (Hashibe et al., 2007). In addition, TP53 codon 72 mutation spectrum has been shown to be altered with Human papillomavirus (HPV) infection, an emerging oral carcinoma FIGURE 1 | Flow diagram of the publication selection process. risk factor (Chor et al., 2016). So far many case-control studies investigated the association of functional polymorphism of TP53 codon 72 with susceptibility to oral carcinoma, but the results remain conflicting and inconclusive (Tandle et al., 2001;Nagpal et al., 2002;Hsieh et al., 2005;Wang et al., 2012;Saleem et al., 2013). We therefore conducted this meta-analyses to evaluate the relationship of TP53 codon 72 polymorphism with tobacco and/or alcohol use and HPV infection in the susceptibility to oral carcinoma.

Search Strategy
PubMed, EMBASE, and Google of Scholar databases up to August 19, 2017 were searched with a combination of the keywords as follows: [(oral OR tongue OR mouth OR buccal OR oropharynx) AND (tumor OR carcinoma OR cancer) AND (TP53 OR P53 OR Arg72Pro) AND (variant * OR mutation OR polymorphism * )].

Inclusion and Exclusion Criteria
Inclusion criteria were: (i) evaluated the association between tobacco and/or alcohol uses, TP53 codon 72 polymorphism, HPV infection, and susceptibility to oral carcinoma; (ii) case-control researches published in English or Chinese; (iii) definite histopathologic diagnosis or clearly reported  the type; (iv) sufficient data to evaluate the ORs and 95%CI, and P-value; (v) genotype distribution was in Hardy-Weinberg equilibrium (HWE). Major exclusion criteria were: (i) Reviews, conference abstracts, case reports; (ii) only-case study; (iii) genotype distribution was inconsistent with HWE; (iv) when duplicated studies published, only the study with the large sample size was included.  Frontiers in Physiology | www.frontiersin.org

Data Extraction
Data extraction was performed independently by two authors using a standardized form. Data such as: first author, country, ethnicity, year of publication, source of the controls, genotype distribution of cases and controls. Discrepancies were settled by discussion, with disagreements resolved by consensus.

Statistical Analysis
The association was determined by calculating odds ratios (ORs) with corresponding 95% credible interval (95%CI). Q-test and I 2 statistics were used to quantify statistical heterogeneity. The random-effect model was conducted if the heterogeneity was significant (P < 0.05) (DerSimonian and Laird, 1986); otherwise, the fixed effect model was utilized (Mantel and Haenszel, 1959). The sensitivity analysis was carried out through sequential exclusion of any one individual study. Begg's funnel plot and the Egger's test was performed to assess the potential publication bias of the researches (Begg and Mazumdar, 1994;Egger et al., 1997). The present meta-analysis was carried out by STATA 12.0 (Stata, College Station, TX, USA). P < 0.05 was considered statistically significant.

Meta-Analysis Results
Based on 21 case-control studies no significant association was found between TP53 Arg72Pro polymorphism and susceptibility to oral carcinoma in any genetic model ( Table 2).
In addition, subgroup analysis stratified by tobacco use (no vs. yes) was performed, and the association was still not significant in either tobacco users (OR = 0.88, 95%CI = 0.67-1.16) or nonusers (OR = 0.84, 95%CI = 0.84-2.26). Similar results were found for subgroup analysis stratified by alcohol use or HPV infection status (Figures 3, 4, Table 3).

Sensitivity and Heterogeneity Analysis
Between-study heterogeneity was examined and significant heterogeneity (P < 0.05) was detected in some genetic comparisons, so random-effects model was adopted (DerSimonian and Laird, 1986); otherwise, the fixed-effect model was utilized (Mantel and Haenszel, 1959). The sensitivity analysis was carried out through sequential exclusion of any one individual study, and the results showed that our conclusion was robust and credible ( Figure 5).

Publication Bias
Begg's and Egger's test was utilized to examine the potential publication bias of the studies (Begg and Mazumdar, 1994;Egger et al., 1997). As shown in Figure 6, there was no significant publication bias (Tables 2, 3).

Discussion
TP53 inactivation is a frequent event in cancer and involves point mutations and allelic loss (Baker et al., 1990;Tommasino et al., 2003). Moreover, TP53 polymorphisms could affect cancer susceptibility (Whibley et al., 2009). Pro72 allele has been implicated in coronary artery disease (Khan et al., 2016), systemic lupus erythematosus (Lee et al., 2012) and ulcerative colitis (Vaji et al., 2011). In contrast, Arg72 allele is implicated in pilocytic astrocytoma (Mascelli et al., 2016). Codon 72 TP53 polymorphisms have shown different associations with the risk of carcinomas in different populations, including oral carcinoma susceptibility (Tandle et al., 2001;Nagpal et al., 2002;Hsieh et al., 2005;Wang et al., 2012Wang et al., , 2014Dahabreh et al., 2013;Saleem et al., 2013). Although many case-control studies investigated the association of tobacco and/or alcohol uses, TP53 codon 72 polymorphism, and HPV infection with oral carcinoma susceptibility, the results were inconclusive. A past case-control studies failed to detect any significant association of TP53 codon 72 polymorphism with oral carcinoma susceptibility (Summersgill et al., 2000). In 2007, Bau et al. reported that the ArgArg genotype seemed to increase the susceptibility to oral carcinoma 2.7-fold in Chinese (Bau et al., 2007). Previous several meta-analyses reported the lack of association between TP53 codon 72 polymorphism and the risk of oral carcinoma (Zhuo et al., 2009b;Zeng et al., 2014;Hou et al., 2015), but these studies did not stratify the conditions such as tobacco and/or alcohol uses, HPV-infection status to perform subgroup analysis. Therefore, we performed the present meta-analysis to provide better estimate on the association of TP53 codon 72 polymorphism with oral carcinoma susceptibility.
Tobacco smoking is a well-known risk factor of cancer and could affect gene polymorphism in oral carcinoma (Ye et al., 2008). To evaluate the association between TP53 polymorphism and oral carcinoma susceptibility in tobacco users, we analyzed all available data extracted from the included studies, and found no significant association, indicating that TP53 codon 72 polymorphism is not a potential risk factor of oral carcinoma in tobacco users.
Regular alcohol consumption is associated with an increased risk for oral cancer. Such association is dose-dependent. Indeed, among individuals consuming 4-5 drinks daily, the risk for cancer of the oral cavity is 2-3-fold higher than among nondrinkers (Baan et al., 2007;Seitz and Stickel, 2007;Wiseman, 2008). To further investigate a possible association between oral carcinoma susceptibility and TP53 codon 72 polymorphism in alcohol users, we extracted relevant data from two studies and the results also failed to suggest a market correlation, demonstrated that TP53 codon 72 polymorphism may not be a risk of oral carcinoma in alcohol use status.
HPV infection has been suggested as one of the contributing factors for oral carcinoma. It was suggested that the interaction of TP53 codon 72 polymorphism with HPV was associated with oral carcinoma susceptibility (Kitkumthorn et al., 2010). However, this opinion was challenged by other studies (Summersgill et al., 2000;Lin et al., 2008). In this study we performed subgroup analysis on the interaction of p53 gene polymorphism with HPV infection on oral cancer susceptibility and the results indicated that TP53 codon 72 polymorphism is not a risk factor of oral carcinoma no matter HPV infection status.
The present study had several limitations. Firstly, only studies written in English or Chinese were included in the meta-analysis. This means that eligible studies published in other languages may have been overlooked, which may have introduced selection bias. Secondly, the sample size of some studies was limited and the results should be interpreted carefully. Thirdly, this study had statistical heterogeneity, although this is extremely common in meta-analyses of genetic association studies. We thus conducted subgroup analyses to identify all factors that contribute to the heterogeneity. Finally, other factors such as the age, gender, life-style that may affect the interaction of TP53 codon 72 polymorphism with oral carcinoma could not be analyzed due to the lack of original data.
In summary, this meta-analysis suggests that there is no statistical association between TP53 codon 72 polymorphism and oral cancer susceptibility, independent of tobacco and/or alcohol use and HPV-infection status. However, this conclusion should be confirmed by multi-center and large-scale studies based on multiple ethnic groups.

AUTHOR CONTRIBUTIONS
Y-ML, JS, X-TZ, and R-HL conceived and designed the experiments. Y-ML performed the experiments. X-HY, CH, and X-WJ analyzed the data. X-WJ, Y-DY, C-JW, E-MZ, and Z-XY contributed reagents, materials, analysis tools. Y-ML wrote the paper. Y-ML, and X-TZ methods analysis.