Molecular Biology Reports

, Volume 38, Issue 8, pp 4847–4853

TP53 codon 72 polymorphism and colorectal cancer susceptibility: a meta-analysis

Authors

  • Jing-Jun Wang
    • Department of Center for Disease Control and Prevention of Shaanxi Province
    • Department of Center for Disease Control and Prevention of Shaanxi Province
  • Liang Sun
    • Department of Center for Disease Control and Prevention of Fuyang
  • Li Wang
    • Department of Center for Disease Control and Prevention of Shaanxi Province
  • Peng-Bo Yu
    • Department of Center for Disease Control and Prevention of Shaanxi Province
  • Jian-Hua Dong
    • Department of Center for Disease Control and Prevention of Shaanxi Province
  • Lei Zhang
    • Department of Center for Disease Control and Prevention of Shaanxi Province
  • Jing Xu
    • Department of Center for Disease Control and Prevention of Shaanxi Province
  • Wei Shi
    • Department of Center for Disease Control and Prevention of Shaanxi Province
  • Yu-Chun Ren
    • Department of Center for Disease Control and Prevention of Shaanxi Province
Article

DOI: 10.1007/s11033-010-0619-8

Cite this article as:
Wang, J., Zheng, Y., Sun, L. et al. Mol Biol Rep (2011) 38: 4847. doi:10.1007/s11033-010-0619-8

Abstract

Colorectal cancer constitutes a significant proportion of the global burden of cancer morbidity and mortality. A number of studies have been conducted to explore whether TP53 codon 72 polymorphism is associated with colorectal cancer susceptibility. However, controversial results were obtained. In order to derive a more precise estimation of the relationship, we systematically searched Medline, Google scholar, and Ovid database for studies reported before May 2010. A total of 3603 colorectal cancer cases and 5524 controls were included. TP53 codon 72 polymorphism was not associated with colorectal cancer risk in all genetic models (for dominant model: OR = 0.99, 95% CI: 0.86–1.15; for recessive model: OR = 1.00, 95% CI: 0.81–1.23; for Arg/Pro vs. Arg/Arg: OR = 1.00, 95% CI: 0.87–1.15; for Pro/Pro vs. Arg/Arg: OR = 0.97, 95% CI: 0.76–1.25). In the subgroup analyses by ethnic groups and sources of controls, no significant associations were found in all models. Taken together, this meta-analysis suggested that the biologically usefulness of TP53 codon 72 polymorphism as a selection marker in colorectal cancer susceptibility may be very limited.

Keywords

TP53Codon 72Colorectal cancerMeta-analysis

Introduction

Colorectal (colon and rectum) cancer constitutes a significant proportion of the global burden of cancer morbidity and mortality [1]. Approximately 1 million new cases are diagnosed, and more than half a million people die from colorectal cancer every year [1]. Although numerous epidemiological and biological studies have revealed risk/protective factors for colorectal cancer, present knowledge is still insufficient to elucidate the etiological mechanisms of the disease [2].

An explosion of information and insights into the molecular pathogenesis of sporadic colorectal cancer can be date back to the late 1980 s, and since then it has served as a paradigm for the investigation of cancer genetics in general [3]. TP53 is one of the most extensively studied genes as a tumor suppressor [4]. It has been thought to plays a pivotal role in modulating cell growth, division, and apoptosis. Mutant TP53 may contribute to increased cell proliferation, loss of ability to undergo apoptosis, and increasing genetic instability [5]. An important TP53 polymorphism is the restriction fragment length polymorphism in codon 72 of exon 4 coding for proline (72Pro: CCC) or arginine (72Arg: CGC) [6]. The both structural forms have been shown to have some different biochemical and biological properties [7], such as different binding to components of the transcriptional machinery and different activation of transcription [8].

Recently, the role of TP53 codon 72 polymorphism in the etiology of different types of cancer has drawn more and more attention, including colorectal cancer. A number of studies have been conducted to explore whether TP53 codon 72 polymorphism is associated with colorectal cancer susceptibility [931]. However, the results of these studies remain conflicting rather than conclusive. We therefore conducted a meta-analysis to more precisely define the effect of TP53 codon 72 polymorphism on risk for colorectal cancer.

Methods

Search strategy and selection criteria

We systematically searched Medline, Google scholar, and Ovid database for studies reported before May 2010, without language restriction, using the search terms: TP53, P53, and colorectal. Non-English articles were translated if necessary. Review articles and bibliographies of relevant literatures were manually scanned to identify eligible studies. Studies were selected according to the following criteria: (a) The study used a case–control study design, (b) the report had available genotype frequency, in the case of the literature without genotype frequency reported, we contact with the author for unavailable genotype frequency, (c) In the case of duplication with multiple articles publishing data on the same population, the most complete data set was included. (d) The gene distributions of control groups were in agreement with Hardy–Weinberg equilibrium (HWE).

Data extraction

Two investigators independently extracted data using a standardized data extraction form. Discrepancies were resolved by discussion and if consensus was not achieved, the decision was made by the third investigator. The title and abstract of all potentially relevant articles were screened to determine the irrelevance. Full articles were also scrutinized if the title and abstract were ambiguous. We extracted standardized data sets from studies of TP53 codon 72 polymorphism and colorectal cancer. The following information was sought from each publication: authors, years of publication, countries of origin, ethnicities of participants, sources of controls, numbers of cases and controls, and study designs. For articles including different populations, data were extracted separately (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs11033-010-0619-8/MediaObjects/11033_2010_619_Fig1_HTML.gif
Fig. 1

Forest plot for the overall association between TP53 codon 72 polymorphism and colorectal cancer risk. s.e., standard error aDominant model Arg/Pro and Pro/Pro vs. Arg/Arg. bRecessive model Pro/Pro vs. Arg/Arg and Arg/Pro cCodominant model Arg/Pro vs. Arg/Arg dCodominant model Pro/Pro vs. Arg/Arg

Data analysis methods

Pooled ORs and 95% CIs were used to assess the strength of the associations. We calculated pooled ORs and 95% CIs for all studies combined. Furthermore, subgroup analyses were performed by ethnic groups and sources of controls. Ethnic group fewer than three studies and mixed population were grouped together as “Mix” in analyses by ethnicities.

For the assessment of the deviation from HWE in the reported genotype frequencies among controls, the appropriate goodness-of-fit Chi-square test was performed [32]. Statistical heterogeneity among studies was estimated by use of the Q and I2 statistic [33]. P value greater than 0.10 for the Q test indicates a lack of heterogeneity among studies. Dependent on the results of heterogeneity test among individual studies, the fixed effect model (Mantel–Haenszel) or random effect model (DerSimonian and Laird) was selected to summarize the Pooled OR. A sensitivity analysis was performed to illustrate the accuracy and stability of the analytic results. Sensitivity analyses were conducted by deleting a single study each time involved in the meta-analysis [34]. Publication bias was investigated with funnel plots, in which the standard error of log OR of each study was plotted against its OR. An asymmetric plot suggested possible publication bias. The significance of the intercept was determined by the method of the Egger’s linear regression test. Furthermore, Begg’s rank correlation test was performed to check the publication bias. P value < 0.05 was considered representative of statistically significant. Stata version10 (Stata Corp, College Station, Texas, USA) was used for the statistical analysis.

Result

Literature search and meta-analysis databases

Based on our search criteria, a total of 35 studies were preliminarily eligible [944]. However, after screening of the full articles, seven studies [8, 3540] were excluded because they were not case–control study, and three studies [4143] were irrelevant. One article [44] was excluded because it was conducted on overlapping populations with other eligible study. Three studies [11] [26, 31] were excluded because the gene distribution of control group deviation from the HWE. Finally, Twenty-one studies (3603 colorectal cancer cases and 5524 controls) were included in the meta-analysis. Study characteristics are summarized in Table 1. There were seven studies of Asians, nine studies of Caucasians, and five studies of mixed population.
Table 1

Characteristics of eligible studies about TP53 codon 72 polymorphism and colorectal cancer risk included in the meta-analysis

First author

Year

Country

Ethnic group

Source of control

Case

Control

Study design

Olschwang[9]

1991

France

Caucasian

PB

71

115

Case–control

Kawajiri[10]

1993

Japan

Asian

PB

84

347

Case–control

Murata[12]

1996

Japan

Asian

HB

115

152

Case–control

Wang[13]

1999

China

Asian

PB

61

140

Case–control

Sayhan[14]

2001

Turkey

Mix

PB

67

76

Case–control

Hamajima[15]

2002

Japan

Asian

HB

147

241

Case–control

Gemignani[16]

2004

Spain

Caucasian

HB

352

316

Case–control

Schneider[17]

2004

Germany

Caucasian

PB

57

85

Case–control

Krüger[18]

2005

Germany

Caucasian

PB

293

245

Case–control

Sotamaa(Finland) [19]

2005

Finland

Caucasian

PB

379

323

Case–control

Sotamaa(USA) [19]

2005

USA

Caucasian 87%

PB

30

118

Case–control

Black 10%

Asian 3.3%

Koushik[20]

2006

USA

Caucasian 94%

PB

442

904

Nest

 

Case–control

Lima[21]

2006

Brazil

Caucasian 71%

HB

100

100

Case–control

Black 29%

Perez[22]

2006

Argentina

Hispanic

PB

53

109

Case–control

Perfumo[23]

2006

Italy

Caucasian

Mix

60

146

Case–control

Tan[24]

2007

Germany

Caucasian

PB

467

563

Case–control

Zhu[25]

2007

China

Asian

PB

345

670

Case–control

Csejtei[27]

2008

Hungary

Caucasian

PB

102

97

Case–control

Cao[28]

2009

Korean

Asian

PB

156

293

Case–control

Mammano[29]

2009

Italy

Caucasian

PB

90

321

Case–control

Mojtahedi[30]

2010

Iran

Asian

PB

132

163

Case–control

PB Population-based, HB hospital-based

Quantitative data synthesis

TP53 codon 72 variation was not associated with colorectal cancer risk in all genetic models (for dominant model: OR = 0.99, 95% CI: 0.86–1.15; for recessive model: OR = 1.00, 95% CI: 0.81–1.23; for Arg/Pro vs. Arg/Arg: OR = 1.00, 95% CI: 0.87-1.15; for Pro/Pro vs. Arg/Arg: OR = 0.97, 95% CI: 0.76–1.25, Fig. 2). In the subgroup analyses by ethnic groups and sources of controls, no significant associations were found in all models. Statistical heterogeneity between studies was observed in a few comparisons (Table 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs11033-010-0619-8/MediaObjects/11033_2010_619_Fig2_HTML.gif
Fig. 2

Funnel plot for the overall association between TP53 codon 72 polymorphism and colorectal cancer risk. s.e., standard error aDominant model Arg/Pro and Pro/Pro vs. Arg/Arg. bRecessive model Pro/Pro vs. Arg/Arg and Arg/Pro cCodominant model Arg/Pro vs. Arg/Arg dCodominant model Pro/Pro vs. Arg/Arg

Table 2

Results of meta-analysis for dominant, recessive and codominant models for TP53 codon 72 polymorphism and colorectal cancer

Study groups

Dominant model

Recessive model

(Arg/Pro and Pro/Pro vs. Arg/Arg)

(Pro/Pro vs. Arg/Arg and Arg/Pro)

OR (95% CI)

Ph

I2(%)

OR (95% CI)

Ph

I2 (%)

Overall

0.99 (0.86–1.15)R

0.001

55.8

1.00 (0.81–1.23)R

0.034

39.9

Asian

1.12 (0.87–1.43)R

0.045

53.3

1.19 (0.83–1.71)R

0.009

64.6

Caucasian

0.98 (0.80–1.21)R

0.043

51.7

0.81 (0.61–1.07)

0.870

0.0

Mix

0.78 (0.52–1.17)R

0.033

62.0

0.91 (0.63–1.31)

0.629

0.0

PB

0.95 (0.79–1.13)R

0.001

60.3

1.08 (0.85–1.37)R

0.036

43.6

HB

1.09 (0.88–1.34)

0.315

15.5

0.74 (0.52–1.06)

0.833

0.0

Study groups

Codominant model

Codominant model

(Arg/Pro vs. Arg/Arg)

(Pro/Pro vs. Arg/Arg)

OR (95% CI)

Ph

I2 (%)

OR (95% CI)

Ph

I2(%)

Overall

1.00 (0.87–1.15)R

0.012

46.6

0.97 (0.76–1.25)R

0.005R

51.2

Asian

1.13 (0.96–1.34)

0.177

32.9

1.23 (0.80–1.90)R

0.003R

69.5

Caucasian

1.02 (0.81–1.28)R

0.027

55.6

0.80 (0.60–1.07)

0.857

0.0

Mix

0.81 (0.55–1.19)R

0.065

54.8

0.87 (0.59–1.26)

0.349

10.0

PB

0.94 (0.80–1.10)R

0.025

46.4

1.02 (0.75–1.38)R

0.002R

58.1

HB

1.16 (0.93–1.44)

0.383

1.9

0.77 (0.53–1.12)

0.777

0.0

R Random-effects model were performed

PhP value of heterogeneity test

Sensitivity analyses

A single study involved in the meta-analysis was deleted each time to reflect the influence of the individual data-set to the pooled OR. The corresponding pooled ORs were not materially altered for all genetic models.

Bias diagnostics

The shapes of the funnel plots did not reveal any evidence of obvious asymmetry in dominant model and codominant model (Arg/Pro vs. Arg/Arg). However, obvious asymmetry was observed in recessive model and codominant model (Pro/Pro vs. Arg/Arg). Similarly, the results of Egger’s test did not show any evidence of asymmetry in dominant model (t = −1.17, P = 0.256) and codominant model (for Arg/Pro vs. Arg/Arg: t = −1.04, P = 0.310), whereas evidence of asymmetry was observed in recessive model (t = −3.23, P = 0.005) and codominant model (for Pro/Pro vs. Arg/Arg: t = −3.26, P = 0.004). Furthermore, the results of Begg’s rank correlation test revealed no significant publication bias in dominant (Z = 1.20, P = 0.230), recessive (Z = 1.40, P = 0.163), and codominant model (for Arg/Pro vs. Arg/Arg: Z = 1.01, P = 0.315; for Arg/Arg vs. Pro/Pro: Z = 1.52, P = 0.127).

Discussion

In the last two decades, and especially in recent years, TP53 is one of the most extensively investigated genes as a tumor suppressor [4551]. Accordingly, a few meta-analyses have been published on this issue. Hu [[52] demonstrated TP53 codon72 polymorphism is associated with breast cancer risk within certain populations or regions. Yan [53] reported that Pro allele might increase the risk of lung cancer under recessive genetic model in adenocarcinoma, Asians, and lung cancer stage I. Recently, pooled analysis [54] provides evidence that there is no association between TP53 codon 72 polymorphism and cervical cancer when the analysis is restricted to methodologically sound studies. More recently, Zhu [55] demonstrated that TP53 codon 72 polymorphism is not associated with prostate cancer risk. In the context of colorectal cancer, one meta-analysis has been published [56]. However, the authors didn’t perform on a recessive model.

This meta-analysis includes information on 3603 colorectal cancer cases and 5524 controls indicates TP53 codon 72 polymorphism was not associated with colorectal cancer risk. This finding may point to the validity and the robustness of the results initially presented [48]. More importantly, we revealed that TP53 codon 72 polymorphism is not associated with colorectal cancer in recessive model.

There were differences of genetic backgrounds and the gene-environment interactions in the etiology among different ethnicities. However, no significant association was found in any subgroup of population. Since the hospital-based studies may have some biases because such controls are not representative of the general population, subgroup analyses have also been performed by sources of controls. Similarly, no significant association was found.

Because of the degree of variability among the study characteristics, the results from meta-analyses should be interpreted with an appropriate degree of caution. However, the sensitivity analysis had been performed to confirm the reliability and stability of this meta-analysis. In addition, our inclusion of non-English language reports was important in minimizing a major potential threat to the validity of any meta-analysis-the related threat of a language bias.

Some limitations of this meta-analysis should be discussed. Publication bias is a major problem in performing meta-analysis because the studies with negative results are more likely not to be published. Although we attempted to identify unpublished work in Google scholar, publication biases have been observed in some comparisons. Another potential limitation was that our results were based on unadjusted estimates. More precise analyses can be conducted if individual data was available, which would allow for the adjustment by other covariates including age, sex, and other factors.

Taken together, this meta-analysis suggested that the biologically usefulness of TP53 codon 72 polymorphism as a selection marker in colorectal cancer susceptibility may be very limited. Future studies should focus on gene–gene and gene–environment interactions. The result may be lead to better, comprehensive understanding of the association between the TP53 codon 72 polymorphism and cancer risk.

Conflict of interest

None.

Copyright information

© Springer Science+Business Media B.V. 2010