Lack of Association Between the CYP 1 A 1 Ile 462 Val Polymorphism and Endometrial Cancer Risk : a Meta-analysis

It is reported that endometrial cancer is the most frequent “estrogen-sensitivemalignancy” in women (Key et al., 1988) and that estrogen and its metabolites are known to be both inducers and promoters of endometrial cancer (Herrington et al., 2001). With the change of lifestyle, people have more and more chance to expose to estrogen. Prolonged estrogen stimulation factors, including stimulation by nulliparity, late menopause, and obesity, have been identified as risk factors for the development of endometrial cancer (La Vecchia et al., 1986; Elwood et al., 1977). People with these high risk factors should greatly care for their behavior and lifestyle. In recent years, the genetic background was shown to be involved in the etiology of endometrial cancers (Auersperg et al., 1998). Estrogens are metabolized by CYP1A1 and converted into catecholestrogens 2-hydroxyestradiol and 4-hydroxyestradiol (Martucci et al., 1993). It is obviously that genes encoding for enzymes involved in estrogen metabolism such as CYP1A1 have been hypothesized to be involved in the etiology of these pathologies (Herrington et al., 2001). The potential effect of CYP1A1 polymorphism on endometrial cancer risk is under the hypothesis that increased exposure to 2-OH estrogen might decrease, and increased exposure to 4-OH estrogen might increase, endometrial cancer risk (Doherty et al., 2005).


Introduction
It is reported that endometrial cancer is the most frequent "estrogen-sensitivemalignancy" in women (Key et al., 1988) and that estrogen and its metabolites are known to be both inducers and promoters of endometrial cancer (Herrington et al., 2001).With the change of lifestyle, people have more and more chance to expose to estrogen.Prolonged estrogen stimulation factors, including stimulation by nulliparity, late menopause, and obesity, have been identified as risk factors for the development of endometrial cancer (La Vecchia et al., 1986;Elwood et al., 1977).People with these high risk factors should greatly care for their behavior and lifestyle.In recent years, the genetic background was shown to be involved in the etiology of endometrial cancers (Auersperg et al., 1998).Estrogens are metabolized by CYP1A1 and converted into catecholestrogens 2-hydroxyestradiol and 4-hydroxyestradiol (Martucci et al., 1993).It is obviously that genes encoding for enzymes involved in estrogen metabolism such as CYP1A1 have been hypothesized to be involved in the etiology of these pathologies (Herrington et al., 2001).The potential effect of CYP1A1 polymorphism on endometrial cancer risk is under the hypothesis that increased exposure to 2-OH estrogen might decrease, and increased exposure to 4-OH estrogen might increase, endometrial cancer risk (Doherty et al., 2005).

Lack of Association Between the CYP1A1 Ile462Val
Polymorphism and Endometrial Cancer Risk: a Meta-analysis Xi-Wen Wang 1 , Tian-Yu Zhong 2 , Yun-Hui Xiong 3 , Hai-Bo Lin 1 , Qing-Yi Liu 1 * So far, three common polymorphisms in CYP1A1 have been identified in white populations.The Ile462Val polymorphism, a 2455A>G transition in exon 7 located near the active site of the enzyme (rs1048943) is one of them (Hayashi et al., 1991).Unfortunately, the conclusions on the association between CYP1A1 Ile462Val polymorphism and endometrial cancer risk from different studies are inconsistent.Some studies reported that no significant association was found between CYP1A1 Ile462Val polymorphism and endometrial cancer risk (Sugawara et al., 2001;McGrath et al., 2007;Ashton et al., 2010).However, other studies concluded that there was a significant association between CYP1A1 Ile462Val polymorphism and endometrial cancer risk (Esteller et al., 1997;Sugawara et al., 2003;Doherty et al., 2005;Seremak-Mrozikiewicz et al., 2005;Esinler et al., 2006;Rebbeck et al., 2006;Hirata et al., 2008;Ashton et al., 2010).
The following meta-analysis was performed for the purpose of precisely estimating the association between CYP1A1 Ile462Val polymorphism and endometrial cancer risk.All studies published about CYP1A1 Ile462Val polymorphism and endometrial cancer risk were searched and summarized.In order to ensure the analysis quality, meta-regression analysis, sensitivity analysis and publication bias analysis were took into account in the study.

Eligible studies searching
PubMed and Embase were searched (up until 2 July 2012, following the search strategy: CYP1A1 AND (polymorphism OR mutation OR variation) AND ("endometrial cancer" or "endometrial carcinoma").All the eligible studies were retrieved, and their bibliographies were checked for other relevant publications.

Inclusion criteria
Eligible study must meet all the following inclusion criteria:

Data extraction
Information used for meta-analysis was evaluated and extracted carefully from all the eligible studies independently by two of the authors according to the inclusion criteria listed above.Disagreement was resolved by discussion between them.If they could not reach a consensus, then another author was consulted for the settlement of dispute and a final decision was made by the majority of the votes.The following data was collected from each study: first author's name, publication year, country, ethnicity, source of control, genotyping methods, confirmation of diagnosis, numbers genotyped of cases and controls, frequency of allele.Different descents were categorized as Caucasian, Asian and other.

Statistical methods
The strength of association between CYP1A1 Ile462Val polymorphism and endometrial cancer risk was measured by OR with 95% CI.  (Cochranl, 1954).A P value greater than 0.05 for the Q-test indicates no heterogeneity among studies, and so the fixed-effects model was used for the meta-analysis (Mantel et al., 1959).Otherwise, the random effects model was used (DerSimonian et al., 1986).Quantification of the heterogeneity was done with the I 2 metric (I 2 = (Q -df)/Q), which is independent of the number of studies in the metaanalysis (Esteller et al., 1997).The I 2 values falls within the range 0-100%, with higher values denoting greater degree of heterogeneity (I 2 = 0-25%, no heterogeneity; I 2 = 25-50%, moderate heterogeneity; I 2 = 50-75%, large heterogeneity; I 2 = 75-100%, extreme heterogeneity) (Cochran, 1954;Higgins et al., 2002;Zintzaras et al., 2005) Meta-regression analyses were performed to explore the source of heterogeneity.Sensitivity analyses were carried out by excluding one study at a time to examine the influence of individual studies on the summary effect estimate.An estimate of potential publication bias was carried out by the funnel plot, in which the standard error of log (OR) of each study was plotted against its log (OR).An asymmetric plot suggests potential publication bias.Funnel plot asymmetry was assessed by the method of Egger's linear regression test, a linear regression approach to measure funnel plot asymmetry on the natural logarithm scale of the OR.The significance of the intercept was determined by the t-test suggested by Egger (P<0.05 suggested existence of statistically significant publication bias) (Egger et al., 1997).
All of the statistical tests used in our meta-analysis were performed by STATA version 10.0 (Stata Corporation, College Station, TX).
Owing to meta-analysis focusing on analysis for published studies, the study was exempt from IRB approval.

Study characteristics
The initial search with the key words and subject terms identified a total of 13 studies.Of these, a total of 9 studies met the inclusion criteria (Esteller et al., 1997;Esteller et al., 1997;Sugawara et al., 2003;Doherty et al., 2005;Seremak-Mrozikiewicz et al., 2005;Esinler et al., 2006;McGrath et al., 2007;Hirata et al., 2008;Ashton et al., 2010).Among them, however, one study (Esteller et al., 1997) was excluded because the same data was repeated in other study (Esteller et al., 1997).Another study (Esinler et al., 2006) was ruled out because of violation of HWE.As a result, 7 studies including 1286 cases and 2111 controls were included in the final meta-analysis.Table 1 lists the studies identified and their main characteristics.There were 6 studies for Caucasians, 1 study for Asians.Genotype distributions in the controls of all studies were in agreement with HWE.Table 2 shows the distribution of    the CYP1A1 Ile462Val genotypes and the allele frequency for endometrial cancer patients and controls.

Quantitative synthesis
Table 3 and Figure 1 displayed the main results of the meta-analysis for CYP1A1 Ile462Val polymorphism.When all the 7 studies were pooled into the meta-analysis, there was no evidence for significant association between CYP1A1 Ile462Val polymorphism and endometrial cancer risk

Sensitivity analyses
The influence of individual studies on the summary effect estimate may be displayed using sensitivity analyses in which the meta-analysis estimates are computed omitting one study at a time.All the results were not obviously altered and did not draw different conclusions.The sensitivity analysis figures for different polymorphism models were not shown.

Publication bias
Begg's test and Egger's test were performed to assess the publication bias of studies.The shapes of the funnel plots and Egger's test results did not reveal any obvious asymmetry in any Ile462Val polymorphism models.Figure 2 demonstrated the funnel plots for OR of all models.Furthermore, the corresponding Egger's test results also did not suggest any publication bias (data not show) .

Discussion
Endometrial cancer is the seventh most frequent malignancy among women worldwide (Barrena et al., 2009).Despite the fact that endometrial cancer is one of the gynaecological cancers that carries good overall prognosis because it is often detected at early stages of disease, the incidence and mortality rates of endometrial cancer have been increasing (Fader et al., 2009).So far, some studies have reported the association between CYP1A1 Ile462Val polymorphism and endometrial cancer risk.Unfortunately, these results remain controversial.For the purpose of obtaining a more accurate measurement of the association, we performed this meta-analysis.Based on the meta-analysis results, we can learn that there is no association between CYP1A1 Ile462Val polymorphism and endometrial cancer risk.
It is a generally acknowledged fact that estrogen is a risk factor for the development of endometrial cancer.So it is easily to be understood that the change of estrogen metabolism may be connected with the endometrial cancer risk.It is well recognized that estrogens are metabolized by CYP1A1 and converted into catecholestrogens 2-hydroxyestradiol and 4-hydroxyestradiol.Some studies have reported that the Ile462Val polymorphisms are associated with CYP1A1 activity and inducibility (Petersen et al., 1991;Cosma et al., 1993;Kawajiri et al., 1993;Drakoulis et al., 1994;Taioli et al., 1995;Kiyohara et al., 1996;Zhang et al., 1996;Schwarz et al., 2000).The frequency of Ile462Val mutation tended to higher in CYP1A1 high expression population.Moreover, many studies claimed that CYP1A1 Ile462Val polymorphisms are associated with endometrial cancer risk.But the conclusions are contrary to each other.Some study revealed that CYP1A1 Ile462Val polymorphism can decrease the risk for the endometrial cancer development (Hirata et al., 1964;Doherty et al., 2005;Seremak-Mrozikiewicz et al., 2005).However, other studies reported that CYP1A1 Ile462Val mutation can increase the risk for the endometrial cancer development (Esteller et al., 1997;Esteller et al., 1997;Esinler et al., 2006;).
Several factors may contribute to discordant findings among individual studies.Small sample size is one of them, which often enhances the chance factor for falsepositive or false-negative results.In meta-analysis, however, the false-positive and false-negative results may neutralize each others as large number of studies are pooled, and the increase of overall statistical power ensure a more accurate estimate of association.This metaanalysis estimated the relationship between Ile462Val polymorphisms and endometrial cancer risk.The summarized result indicated that no associations between CYP1A1 Ile462Val polymorphism and endometrial cancer risk were found.Regardless of dominant model, recessive model or co-dominant models, the result is invariably negative.The meta-analysis results are consistent with some studies (Sugawara et al., 2003;McGrath et al., 2007;Ashton et al., 2010).
Heterogeneity is a potential problem that may affect the interpretation of the results.Significant between-study heterogeneity existed in Val vs Ile allele model and (Ile.Val + Val.Val) vs Ile.Ile polymorphism model.When we performed meta-regression analysis in which source of controls and ethnicity were adopted as independent variables, the heterogeneity among studies was effectively removed.It suggested that source of controls is the main factor contributing to potential heterogeneity.The reason may be that the hospital-based case-control studies have inherent selection biases due to the fact that such controls did not represent the general population very well.Such selection biases may distort the results and lead to significant heterogeneity between hospital-based studies.Therefore, using proper and representative populationbased controls is very important to reduce biases in genetic association studies.Some limitations of this meta-analysis must be acknowledged.First, the number of studies and population contained in this meta-analysis was relatively small.To a certain extent, it influences the statistical power.Second, heterogeneity among studies was found in Val vs Ile allele model and (Ile.Val + Val.Val) vs Ile.Ile genotype model.For the sake of reasonable interpretation for results, the exploration of heterogeneity must be done.
In spite of limitations, some advantages may be found in this meta-analysis.First, it performed meta-analysis in different polymorphism models respectively.So that we can examine the relationship between CYP1A1 Ile462Val polymorphisms and endometrial cancer risk in comprehensive way.Second, meta-regression analysis was used to explore the source of heterogeneity.It enables us to better reasonably explain the results.Third, Begg's and Egger's tests did not detect any publication bias, indicating that our results should be unbiased.Fourth, when sensitivity analysis was performed to check the influence of individual studies on the summary effect estimate by omitting one study at a time in all Ile462Val polymorphism models, any single study didn't have significant change on overall result.It indicated that the meta-analysis results were robust.
Based on the limits of the study, future investigation about association between CPY1B1 Ieu432Val polymorphism and endometrial cancer risk should pay more attention to homogeneity among studies and comparability between patients and control.In addition, addressing gene-gene and gene-environment interactions is also imperative.
In conclusion, this meta-analysis involving 7 studies and 3397 subjects strongly suggests that CYP1A1 Ile462Val polymorphism is not associated with endometrial cancer risk.
(a) case-control study evaluating the CYP1A1 Ile462Val polymorphism and endometrial cancer risk; (b) listing the frequency of case and control according to different genotype; (c) full text articles; (d) literature published in English.(e) genotype distribution in the control of the study was in agreement with Hardy-Weinberg equilibrium (HWE).

The pooled OR was estimated for codominant model: Val.Val vs Ile.Ile, dominant model
: (Ile.Val + Val.Val) vs Ile/Ile, and recessive model: Val.Val vs (Ile.Ile + Ile.Val) respectively.Heterogeneity assumption was checked by the chi-square-based Q-test

Table 3 . Results of Meta-analysis for Various Genetic Contrasts of CYP1A1 Ile462Val Polymorphism
r, Fixed effects; *, random effects