Meta-analysis of Associations between the MDM 2T 309 G Polymorphism and Prostate Cancer Risk

Prostate cancer is one of the most commonly diagnosed male malignancies, and it remains a leading cause of death in most Western countries, especially in elderly men (Detchokul et al., 2011). P53 is a well-known tumor suppressor gene which is negatively regulated by the mouse double minute 2 (MDM2) gene. MDM2 could bind to p53 with high affinity, resulting in down-regulation of p53 (Bouska et al., 2009; Eischen a et al., 2009; Kruse et al., 2009). A single-nucleotide polymorphism (SNP-T309G) in the promoter region of MDM2 was proved to increase the expression of MDM2, leading to the attenuation of p53 and increased risk of tumorigenesis (Bond et al., 2004; Vassilev et al., 2004; Bond et al., 2005; Liu et al., 2011; Ma et al., 2012). It is reported MDM2-SNP T309G is associated with increased susceptibility to gastric, cervical and liver cancer (Ohmiya et al., 2006; Yoon et al., 2008; Nunobiki et al., 2010). Independent studies have focused on the association between this polymorphism and prostate cancer risk, but their results were inconclusive, which put forward the requirement of a more comprehensive and


Introduction
Prostate cancer is one of the most commonly diagnosed male malignancies, and it remains a leading cause of death in most Western countries, especially in elderly men (Detchokul et al., 2011). P53 is a well-known tumor suppressor gene which is negatively regulated by the mouse double minute 2 (MDM2) gene. MDM2 could bind to p53 with high affinity, resulting in down-regulation of p53 (Bouska et al., 2009;Eischen a et al., 2009;Kruse et al., 2009).
A single-nucleotide polymorphism (SNP-T309G) in the promoter region of MDM2 was proved to increase the expression of MDM2, leading to the attenuation of p53 and increased risk of tumorigenesis (Bond et al., 2004;Vassilev et al., 2004;Bond et al., 2005;Liu et al., 2011;Ma et al., 2012). It is reported MDM2-SNP T309G is associated with increased susceptibility to gastric, cervical and liver cancer (Ohmiya et al., 2006;Yoon et al., 2008;Nunobiki et al., 2010). Independent studies have focused on the association between this polymorphism and prostate cancer risk, but their results were inconclusive, which put forward the requirement of a more comprehensive and

Meta-analysis of Associations between the MDM2-T309G Polymorphism and Prostate Cancer Risk
Tao Chen 1 , Shang-Hui Yi 2 , Xiao-Yu Liu 1 , Zhi-Gang Liu 1 * reliable assessment of the polymorphism and prostate cancer risk.
Here, we performed a meta-analysis of 4 casecontrol studies (732 cases and 836 controls) to explore their relationship. During which we found that the polymorphism of Tg G change in -309 of MDM2 promoter may have no significant overall effect on the risk of prostate cancer, while it probably can be a protective factor to prostate cancer in European population and hospital-based population, though larger number of samples are required to clarify in future.

Selection of studies and data collection
We performed a systematic search of literature prior to July 2012 from PubMed and Medline with the terms of "prostate cancer" and "MDM2", resulting in 112 eligible publication candidates. After comprehensive screening, 4 independent studies published from 2008 to 2010 were selected for the meta-analysis, and the references of the 4 were also screened. Specific search workflow is shown in Figure 1. Then the following basic data was extracted from  those studies for further analysis: first author, published year, source of controls, location, population (ethnicity), sample size and genotype distributions.

Statistical analysis
We performed overall as well as subgroup metaanalysis stratified by ethnicity or source of controls. Pooled odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for evaluating the association between MDM2-SNP T309G and prostate cancer risk. We applied 4 genetic models (GT/TT, GG/TT, dominant model and recessive model) and 2 mathematical models (fixed effects model and random effects model) to calculate in order to get a more comprehensive analysis. Heterogeneity test was used to determine which mathematical model is more suitable. When the chi-square-based Q-test resulted a P value more than 0.10, fixed effects model was considered to be more precise, otherwise, random effects model was used. Egger's regression test was used to test publication bias and a P value more than 0.05 was considered as absence of publication bias.
In this study, we used R software (version 2.12.1) and the Meta package for R (www.r-project.org) to conduct all the analysis.

Study characteristics
4 case-control studies (Kibel et al., 2008;Stoehr et al., 2008;Mandal et al., 2010;Xu et al., 2010), including 732 cases and 836 controls, published from 2008 to 2010 investigated the association between MDM2-SNP T309G and prostate cancer risk. The main characteristics of the four studies were shown in Table.1. Two of the studies carried out their research in Asian population while one in European population, the rest one by Adam was conducted in St.Louis, Missouri, USA, but they reported those subjects were all European descendants, thus, in the subgroup analysis, we considered them as European. Three of the studies recruited controls from hospital, which was used as another stratification criterion in subgroup analysis.

Discussion
MDM2 could bind to the N-terminal transactivation domain of p53, which functions as the principal endogenous E3-ligase with high specificity, thereby negatively modulates its transcriptional activity and stability (Bouska et al., 2009;Eischen et al., 2009;Kruse et al., 2009). SNP T309G, found in the MDM2 promoter, could increase the affinity of the transcriptional activator Sp1, leading to increased expression of MDM2 the subsequent down-regulation of p53 (Bond et al., 2004). It is reported the MDM2-SNP T309G is associated with not only increased risk of gastric carcinoma, cervical cancer and liver cancer but also poor prognosis of several cancer types (Ohmiya et al., 2006). However, some studies showed absence of significant effect of SNP T309G on tumorigenesis. Brenda et al found the polymorphism was not statistically associated with breast cancer risk among African American or Caucasian women (Boersma et al., 2006), while data from a Chinese group showed no close connection between SNP T309G and breast cancer either (Ma et al., 2006).
Four independent studies were involved in the relationship between MDM2-SNP T309G and prostate cancer risk. Two of them (Kibel et al., 2008;Mandal et al., 2010) suggested the polymorphism acted as riskreduced factor in developing prostate cancer and the rest two found no significant association. Besides, Jaboin reported the polymorphism was not associated with clinicopathologic variables, recurrence risk, and overall survival outcome in prostate cancer (Jaboin et al., 2011). Our meta-analysis of them revealed, overall, MDM2-SNP T309G did not increase the susceptibility to prostate cancer, while in European population, two genetic models found MDM2-SNP T309G to be a protective factor against prostate cancer which was different from that of Asian population. It indicated that the polymorphism probably played different role in different populations and compensatory mechanism was likely to exist in the P53 pathway, making the relationship between MDM2-SNP T309G and prostate cancer distinct from other cancer types. Meanwhile, another subgroup analysis considered the polymorphism to reduce prostate cancer risk in hospital-recruited population, and the result needed careful interpretation because of the relative small pooled sample size. However, a study carried out in a more reliable sample size of Caucasian population (Sun et al., 2010) could partially support our results, they found the Mdm2 SNP309 T allele was associated with earlier onset prostate cancer (P = 0.004), higher Gleason scores (P = 0.004), and higher stages in men undergoing a radical prostatectomy (P = 0.011).
In summary, this analysis provided a more comprehensive understanding of relation between MDM2-SNP T309G and prostate cancer risk. We found the TgG change in +309 of MDM2 promoter have no significant overall effect on the risk of prostate cancer, however, the polymorphism probably can be a protective factor to prostate cancer in European population and hospitalbased population despite a larger number of samples are required to testify.