Murine double minute 2 promoter SNP309 polymorphism and prostate cancer risk: A meta-analysis

Authors


Long Yu M.D., Ph.D., The State Key Laboratory of Genetic Engineering, Fudan University, 220 Handan Road, Shanghai 200433, China. Email: longyu@fudan.edu.cn

Abstract

Objective:  The murine double minute 2 gene encodes a negative regulator of the tumor protein p53. A single nucleotide polymorphism in murine double minute 2 promoter, SNP309 T>G, has been reported to alter murine double minute 2 protein expression and to accelerate tumor formation in humans. We carried out a meta-analysis to explore the association between this polymorphism and prostate cancer risk.

Methods:  All eligible studies were searched in PubMed. Crude odds ratios, with 95% confidence intervals, were assessed for the association using fixed- and random-effects models.

Results:  Overall, five case–control studies (872 cases, 1005 controls) were included in the meta-analysis. A significant association between murine double minute 2 SNP309 and prostate cancer risk was observed for homozygote genetic model GG versus TT (odds ratio 0.72, 95% confidence interval 0.55–0.95, P < 0.05, P = 0.130 for heterogeneity), and for dominant model TG + GG versus TT (odds ratio 0.79, 95% confidence interval 0.65–0.96, P < 0.05, P = 0.119 for heterogeneity). The stratified analysis based on ethnicity showed a significant effect of the polymorphism on prostate cancer risk in Caucasians for GG versus TT.

Conclusions:  Findings of the present meta-analysis suggest that the murine double minute 2 309 G allele might be associated with a reduced risk of prostate cancer. The effect of murine double minute 2 309 G allele on tumorigenesis might be influenced by sex and hormonal status.

Abbreviations & Acronyms
CI =

confidence interval

HWE =

Hardy–Weinberg equilibrium

MDM2 =

murine double minute 2

OR =

odds ratio

PCR–RFLP =

polymerase chain reaction restriction fragment length polymorphism

SNP =

single nucleotide polymorphism

Introduction

Prostate cancer is the most common non-skin cancer diagnosed amongst North American and European men. The incidence of prostate cancer has increased rapidly over the past two decades, and it is the second and third most common cause of cancer death in the USA and Europe, respectively.1,2 Various risk factors are associated with prostate cancer, including diet,3 lifestyle,4 hormones,5 family history6 and so on. Despite its high morbidity, the pathogenesis of the development and progression of prostate cancer is far from being clear at present, and genetic factors are considered to contribute substantially to the development of sporadic prostate cancer.7 Dysregulation of cell cycle pathway genes, such as tumor suppressor gene P53 mutations and inactivation, have been reported to play an important role during multistep of prostate cancer carcinogenesis.8

The MDM2 gene encodes a direct negative regulator for the p53 protein.9MDM2 overexpression is associated with advanced tumor stage, and increased cell proliferation and tumor volume of prostate cancer.10,11MDM2 antagonists in combination with androgen depletion could lead to a greater tumor regression and dramatically increased survival rate of prostate cancer cell-bearing nude mice.12 A SNP in the promoter region of MDM2, SNP309 T>G (a change from T to G, rs2279744), is a functionally significant SNP in the promoter region of MDM2, which occurs at nucleotide 309 in intron 1 of the MDM2 gene.13 The polymorphism is suggested to be associated with the risk and early onset age of various human cancers.14,15 In recent years, several studies have investigated the association between MDM2 SNP309T>G polymorphism and prostate cancer risk.16–20 However, the results remain inconsistent and inconclusive, partially because of the relatively small sample size of individual studies and sampling effects. Therefore, we carried out a meta-analysis to obtain a more precise estimation of the association.

Methods

Publication search

We searched for studies using the following search terms: “MDM2”, and “SNP”, “polymorphism”, “mutation” or “variant” and “prostate cancer” in Medline database utilizing the PubMed engine (updated on 20 December 2011). We evaluated all associated publications to identify the most eligible literature. Their reference lists were hand-searched to obtain other relevant publications. The results were limited to English language papers.

Inclusion and exclusion criteria

The following criteria were used to select literature for further meta-analysis: (i) published in peer-reviewed journals; (ii) articles about MDM2 SNP309T>G polymorphism and risk of prostate cancer; and (iii) containing useful genotype frequencies. The major exclusion criteria were: (i) not case–control studies; (ii) control population including malignant tumor patients; and (iii) duplication of a previous publication.

Data extraction

Two investigators (Liu and Jiang) reviewed and extracted information from all eligible publications independently, according to the aforementioned inclusion and exclusion criteria. In the case of settling conflicts, an agreement was reached by discussion between reviewers. The following characteristics were collected from each study: first author's surname, year of publication, country of origin, ethnicity, source of control and case groups, specimen of cases, genotyping methods for assessment of MDM2 SNP309T>G genotypes, total number of cases and controls, as well as number of cases and controls with T/T, T/G and G/G genotypes, respectively.

Quality score assessment

The quality of studies was also independently assessed by the same two reviewers according to the predefined scale of quality assessment (Table 1). These scores were determined using both traditional epidemiological consideration and cancer genetic issues. Any disagreement was resolved by discussion between reviewers. The total score ranged from 0 (worst) to 15 (best). Reports scoring <10 were classified as “low quality” and those scoring ≥10 as “high quality.”

Table 1. Scale for quality assessment
CriteriaScore
Source of cases 
 Selected from population or cancer registry3
 Selected from hospital2
 Selected from pathology archives, but without description1
 Not described0
Source of controls 
 Population-based3
 Blood donors or volunteers2
 Hospital-based (cancer-free patients)1
 Not described0
Specimens of cases determining genotypes 
 White blood cells or normal tissues3
 Tumor tissues or exfoliated cells of tissue0
HWE in controls 
 HWE3
 Hardy–Weinberg disequilibrium0
Total sample size 
 >10003
 >500 but <10002
 >200 but <5001
 <2000

Statistic analysis

First, we assessed the HWE for each study using the goodness-of-fit test (χ2 of Fisher's exact test) only in control groups. Crude ORs with 95% CI were used to assess the strength of association between MDM2 SNP309T>G polymorphism and prostate cancer risk. The pooled ORs were carried out for the co-dominant model (homozygote comparison: GG vs TT, heterozygote comparison: GG vs TG), dominant model (TG + GG vs TT) and recessive model (GG vs TT + TG), respectively. Stratified analyses were carried out by sample size (cases and controls in total: ≥400 and <400) and ethnicity (Asian and Caucasian). A χ2-based Q-test was carried out to check the heterogeneity of the ORs.21 If the result of the heterogeneity test was Pheterogeneity ≥ 0.1, ORs were pooled according to the fixed-effects model (Mantel–Haenszel model).22 Otherwise, the random-effects model (DerSimonian and Laird model) was used.23 One-way sensitivity analyses were carried out to access the stability of the meta-analysis' results.24 The potential publication bias was estimated using Egger's linear regression test by visual inspection of the funnel plot. Pheterogeneity < 0.05 was considered representative of statistically significant publication bias.25 If publication bias existed, the Duval and Tweedie non-parametric “trim and fill” method was used to adjust for it.26 All statistical tests were carried out with stata version 10.0 (Stata, College Station, TX, USA).

Results

Study characteristics

We identified 106 potentially relevant articles from our search of the published literature, of which 95 articles were excluded, and 11 studies were identified through the literature search and selection based on the inclusion criteria. During the extraction of data, we excluded five articles27–31 that were not case–control studies and one review article.32 Therefore, five case–control studies including 872 prostate cancer cases and 1005 controls (Published from 2008 to 2010) were identified and included in the final meta-analysis (Fig. 1). The main characteristics of these studies are presented in Table 2. Sample size of the five studies ranged from 269 to 477. Of the five studies, three were of Asians,18–20 and two of Caucasian16,17 (Stoehr et al. provided the ethnicity information of their study17 by pers. comm.). The quality scores for the individual studies ranged from 10 to 11, and all of the studies were classified as high quality. The distribution of the MDM2 SNP309 genotype frequencies among prostate cancer cases and controls of the five studies are listed in Table 3, and the genotype distribution in the controls of all included studies were consistent with HWE.

Figure 1.

Flow diagram of article selection.

Table 2. Main characteristics of studies included in the meta-analysis
First author (reference)YearCountryEthnicitySample source (cases/controls)Specimen of casesSample size (case/control)Genotyping methodQuality score
Kibel162008AmericaCaucasianHospital/hospitalWhite blood cell186/222Pyrosequencing10
Stoehr172008GermanCaucasianHospital/hospitalNormal prostate/White blood cell145/124PCR-RFLP10
Hirata182009JapanAsianHospital/VolunteersWhite blood cell140/167PCR-RFLP11
Xu192010ChinaAsianHospital/hospitalWhite blood cell209/268PCR-RFLP10
Mandal202010IndiaAsianHospital/hospitalWhite blood cell192/224PCR-RFLP10
Table 3. Distribution of MDM2 promoter SNP309 T>G genotype among prostate cancer cases and controls included in the meta-analysis
First author (reference)Genotype P-value of HWE
CasesControls
T/TT/GG/GT/TT/GG/G
n % n % n % n % n % n %CasesControls
Kibel16854688471379041984532150.1230.529
Stoehr176142.16645.51812.44133.16451.61915.30.9820.463
Hirata185841564026195634794732190.0650.661
Xu19442111856.54722.56825.414353.35721.30.0610.259
Mandal206734.97137.05428.15323.79843.87332.6<0.0010.077

Meta-analysis results

Table 4 shows the detailed results of the association between MDM2 SNP309T>G polymorphism and prostate cancer risk, and the heterogeneity test. In the overall analysis, significant associations were observed in the homozygote genetic model GG versus TT (OR 0.72, 95% CI 0.55–0.95, P = 0.018, Pheterogeneity = 0.130; Fig. 2), and for the dominant model TG + GG versus TT (OR 0.79, 95% CI 0.65–0.96, P = 0.020, Pheterogeneity = 0.119), but not in other genetic models (OR 0.91, 95% CI 0.71–1.17 for the heterozygote genetic model GG vs TG, P = 0.480, Pheterogeneity = 0.321; OR 0.83, 95% CI 0.66–1.05 for the recessive model GG vs TT + TG, P = 0.119, Pheterogeneity = 0.290). No significant association was observed in the stratified analysis based on sample size (<400 or ≥400). In the subgroup analysis based on ethnicity, the significant effect of this polymorphism on prostate cancer risk in the Caucasian population was observed in the homozygote model GG versus TT (OR 0.51, 95% CI 0.31–0.86, P = 0.011, Pheterogeneity = 0.458; Fig. 2), as well as in the recessive model GG versus TT + TG (OR 0.58, 95% CI 0.36–0.93, P = 0.024, Pheterogeneity = 0.246), but not in the heterozygote genetic and dominant models. However, no significant association was observed in any genetic model in the Asian population (Fig. 2).

Table 4. Results of meta-analysis for MDM2 promoter SNP309 T>G and prostate cancer risk
Study groups n Sample size (case/control)GG vs TTGG vs TGTG+GG vs TTGG vs TT+TG
OR (95% CI) P P § OR (95% CI) P P § OR (95% CI) P P § OR (95% CI) P P §
  1. †Number of comparisons. ‡P-value for the association. §P-value for heterogeneity.

Total5872/10050.72 (0.55–0.95)0.0180.1300.91 (0.71–1.17)0.4800.3210.79 (0.65–0.96)0.0200.1190.83 (0.66–1.05)0.1190.290
Sample size              
 ≥4003587/7140.72 (0.52–1.00)0.0500.0310.87 (0.65–1.16)0.3470.1280.84 (0.67–1.07)0.1660.0380.81 (0.62–1.07)0.1350.096
 <4002285/2910.72 (0.44–1.77)0.1860.6791.04 (0.65–1.68)0.8570.6510.70 (0.50–0.98)0.0380.8920.89 (0.57–1.38)0.5890.655
Ethnicity              
 Asian3541/6590.83 (0.60–1.13)0.2360.1771.04 (0.78–1.39)0.8010.9370.81 (0.63–1.05)0.1060.0310.93 (0.71–1.22)0.6200.655
 Caucasian2340/3460.51 (0.31–0.86)0.0110.4580.63 (0.38–1.04)0.0710.1720.76 (0.56–1.04)0.0880.5580.58 (0.36–0.93)0.0240.246
Figure 2.

Forest plots of the relationship between MDM2 SNP309T>G polymorphism and prostate cancer risk in homozygote comparison (GG vs TT) by ethnicity. The size of the black square corresponding to each study is proportional to the sample size and thus the weight used in the meta-analysis, and the center of each square represents the OR. The horizontal line shows the corresponding 95% CI of the OR. The pooled OR was obtained using a fixed-effects model and is represented by hollow diamonds, where its center indicates the OR, and its ends correspond to the 95% CI. The weighting factors (weight %) used to calculate the aggregate odds ratio, calculated from the inverse of the variance, is given for each study.

Sensitivity analyses

A single study involved in the meta-analysis was deleted each time to reflect the influence of individual data-set to the pooled ORs, and most of the corresponding pooled ORs were not materially altered (data not shown).

Publication bias

Both Begg's funnel plot and Egger's test were carried out to assess the publication bias of the literature. Begg's funnel plots did not show any evidence of obvious asymmetry in any genetic model (figures not shown). Then, the Egger's test was used to provide statistical evidence of funnel plot symmetry. The results still did not suggest any obvious evidence of publication bias for all genetic models (P = 0.538 for homozygote comparison, P = 0.359 for heterozygote comparison, P = 0.692 for dominant model and P = 0.300 for recessive model).

Discussion

Prostate cancer carcinogenesis is a complex process associated with an accumulation of genetic and epigenetic changes that occur during initiation, promotion and progression of the disease. Since the identification of MDM2 SNP309T>G,13 several studies16–20 have investigated the genetic effects of MDM2 SNP309 on prostate cancer risk. Among these studies, two studies16,20 involving a European-American and an Indian population have shown that the G allele of MDM2 SNP309 is associated with a reduced prostate cancer risk; three studies showed no evidence of an association between this polymorphism and prostate cancer risk, including two studies18,19 involving Japanese and Chinese populations, and one study17 involving a German population. In the present study, we carried out a meta-analysis with the five published studies including 872 cases and 1005 controls to explore the overall effects of MDM2 SNP309 on prostate cancer risk. The main finding of this meta-analysis was that there is a significant association between MDM2 SNP309T>G polymorphism and prostate cancer risk, and the G allele is a reduced risk factor of prostate cancer. No effect of this polymorphism on prostate cancer might be a result of methodological errors, such as selection bias, or limited statistical power.

Although previous cancer-related meta-analyses observed that the G allele of MDM2 SNP309 was a high risk allele in various cancers, such as colorectal cancer,33 breast cancer,34 lung cancer,35 hepatocellular carcinoma36 and so on, the present study showed that the G allele is low-risk allele in prostate cancer. We consider the possible explanation for the conflicting observation is that the effect of the MDM2 SNP309 G allele on tumorigenesis might be influenced by sex and hormones.37,38 Bond et al. showed that the G allele has a higher affinity binding of transcription factor SP1 (a coactivator for many hormone receptors) to the MDM2 promoter region, and an increase in MDM2 gene transcription and subsequent attenuation of p53 pathway-mediated tumor suppression.13 Although the estrogen receptor also binds to the same region,13,36 and G allele of MDM2 was shown to accelerate tumorigenesis and increase cancer risk in women, but not in men, for colorectal cancer, lung cancer and for highly estrogen receptor positive, but not for estrogen receptor negative, breast invasive ductal carcinoma.37–39 Hu et al. provided a model showing that estrogen preferentially stimulated transcription of the MDM2 gene with the G allele of SNP309, and women with a GG genotype for MDM2 could be affected differently by estrogen than women with a TT genotype.40 Having considered the great importance of the androgen axis in prostate carcinogenesis,5 we suspect MDM2 SNP309 might also be affected by hormonal signaling in prostate cancer development. Sun et al. analyzed the association of MDM2 SNP309 with clinicopathological variables in a large hospital-based Caucasian prostate cancer cohort (n = 4073), and found that the MDM2 SNP309 T allele was associated with earlier onset, higher Gleason scores and higher stages in patients undergoing radical prostatectomy.27 These findings are consistent with the present findings that the T allele is a high prostate cancer risk. Thus, MDM2 SNP309 might contribute not only to prostate cancer development, but also to its progression, it could be used as a risk predictor and molecular marker for therapy for prostate cancer in the future. The mechanism of MDM2 SNP309 on prostate cancer tumorigenesis requires further exploration.

Additionally, our subgroup meta-analysis based on ethnicity showed that MDM2 SNP309 T>G polymorphism can significantly influence prostate cancer risk in Caucasians, but not in Asians. It suggests that ethnicity might influence the role of MDM2 SNP309 polymorphism in prostate cancer susceptibility. However, we should be very cautious to draw the conclusion, as just two studies involving Caucasian populations were available for analysis.16,17 More studies with larger samples are warranted.

Some limitations of the present meta-analysis should be addressed. First, all the eligible studies were published papers written in English cited by PubMed. It is possible that some relevant unpublished studies written in English or published papers in other languages that might meet the inclusion criteria were missed. Thus, some inevitable publication bias might exist in the results, although the funnel plots, as well as Egger's linear regression tests, showed no remarkable publication bias in the meta-analysis. Second, just five eligible studies16–20 were analyzed, and two studies17,18 were carried out with a small sample size of <400 in total, which is too small to have enough statistical power to find the real association. Third, in the subgroup analyses by ethnicity, just two studies16,17 were carried out in Caucasians as aforementioned. Therefore, to carry out a more precise analysis of this functional polymorphism on prostate cancer risk, additional studies with larger sample sizes and different ethnicities, especially Caucasians and Africans (not eligible presently for this analysis), are warranted.

Despite the limitations, the results of this meta-analysis suggest that the MDM2 309 G allele significantly decreased the risk of developing prostate cancer. This finding supports the hypothesis that genetic variability in the p53 pathway might contribute to the pathogenesis of prostate cancer, and MDM2 SNP309 might also be affected by hormonal signaling in prostate cancer development. Further well-designed studies with large sample sizes are required to confirm the present findings, and a comprehensive understanding of the association would be helpful in the future field of individualized preventive care.

Acknowledgments

This work was supported by the National 973 program of China (2004CB518605), the National 863 project of China (2006AA020501), the National Key Sci-Tech Special Project of China (2008ZX10002-020), the Project of the Shanghai Municipal Science and Technology Commission (03dz14086) and the National Natural Science Foundation of China (30024001, 30771188).

Conflict of interest

None declared.

Ancillary