Methylenetetrahydrofolate reductase genetic polymorphisms and esophageal squamous cell carcinoma susceptibility: A meta-analysis of case-control studies

Objectives: Genetic factors and environmental factors play a role in pathogenesis of esophageal squamous cell carcinoma (ESCC). Previous studies regarding the association of folate intake and Methylenetetrahydrofolate reductase C677T polymorphism with ESCC was conflicting. We conducted a meta-analysis to investigate the association of MTHFR C677T and folate intake with esophageal cancer risk. Methodology: MEDLINE, EMBASE and the Chinese Biomedical Database were searched in our study. The quality of studies were evaluated by predefined scale, and The association of polymorphisms of MTHFR C677T and folate intake and ESCC risk was estimated by Odds ratio (ORs) with 95% confidence intervals (CIs). Results: Nineteen studies (4239 cases and 5575 controls) were included for meta-analysis. A significant association was seen between individuals with MTHFR 677 CT [OR(95%)=1.47(1.32-1.63)] and TT [OR(95%)=1.69(1.49-1.91)] genotypes and ESCC risk (p<0.05). Low intake of folate had significantly higher risk of esophageal cancer among individuals with CT/TT genotype [OR(95%)=1.65(1.1-2.49)], while high intake of folate did not find significant high risk of esophageal cancer among individuals with CT/TT genotype [OR(95%)=1.64 (0.82-3.26)]. Conclusions: Our meta-analysis indicated the folate intake and MTHFR 677CT/TT are associated with the risk of ESCC, and folate showed a significant interaction with polymorphism of MTHFR C677T.


INTRODUCTION
Esophageal cancer has been ranked as one of the most common malignancies worldwide. 1 Comparing with Comparing with esophageal adenocarcinoma which is one of the most prevalent cancer in Western Countries, and the major phenotype in Asian countries, especially in China, is esophageal squamous cell carcinoma(ESCC). In China, and it is estimated there are new diagnosed 250,000 cases every year. Possible risk factors for ESCC include smoking and drinking habits, high intake of hot-temperature food, salted food and pickled vegetables, low intake of vegetable, chronic mucosal irritation. [2][3][4][5] It is also reported that deficiency of nutrients, such as vitamins and microelements, is association with an enhanced risk for ESCC. 5 It is well known folate is a water-soluble vitamin element from green leafy vegetables, cereals, legumes and fruits. 6 Lack of folate may damage the function of DNA repair and chromosomal fragile site expression, and induce the breaks of chromosome and formation of micronucleus. 6 Folate is also function as a major methyl group donor and had a key role in DNA methylation. 7 Previous studies have shown the aberrant DNA methylation is associated with cancer risk. 7 The common polymorphism for Methylenetetrahydrofolate reductase (MTHFR) gene is a C to T substitution at nucleotide 677 (MTHFR C677T), and this variation leads to a reduced activity of this enzyme. Thus the variation of folate metabolic enzyme activity altered by the genetic polymorphisms may have a role in the susceptibility to methylation-related carcinogenesis. However, the role of dietary folate in esophageal cancer and its relationship between MTHFR C677T were controversial. [8][9][10] The inconsistency of these results may be explained by differences in source of selected subjects, sample size, study design and random error. We therefore performed a systematic review to investigate potential genetic associations of MTHFR C677T and folate intake with ESCC risk by reducing random error and obtaining precise estimates. 11

METHODOLOGY
Searching strategy: We searched PubMed (from Jan. 1966 to Oct. 2012), EMBASE (from January 1988 to Oct. 2012), and the Chinese Journal Net (CNKI; from January 1980 to Oct. 2012) by using the following key words for searching published papers: 'esophageal squamous cell carcinoma', 'esophagus' or 'oesophagus', 'carcinoma or cancer or neoplasm or tumour or tumor', 'Methylenetetrahydrofolate reductase', or 'MTHFR'. We only extract published paper in English and Chinese. All references cited in studies and published review articles were extracted for additional eligible studies. The criteria of including data were as follows: (1) a case-control study design; (2) reporting the association between MTHFR C677T polymorphisms and ESCC; (3) original data about the genotype distributions of MTHFR C677T genotypes. (4) If the data were published for more than twice, we only extracted the most complete data. Two reviewers independently evaluated and extracted the potential selected articles, and the disagreements were resolved by discussion. If the data were missing, we attempted to contact the authors of the articles by email or phone so as to request the relevant data. All the data were organized as first author's name, year of publication, study site, numbers of cases and controls, genotype frequencies of MTHFR C677T.
The quality of studies was evaluated by a previous defined scale reported by Jiang (Table-I). 12 The total quality score was ranged from 0 to 15. Score≤10 was regarded as low quality of study, and score >10 was regarded as high quality of study.
Statistical analysis: All statistical analyses were performed using STATA statistical package (Version 9.1, STATA, College Station, TX). Variation of genotype frequency distribution in controls from the expected under Hardy-Weinberg equilibrium (HWE) were calculated using Chi-square test. The pooled Odds ratio (ORs) and 95% confidence intervals (CIs) were used to assess the association of polymorphisms in MTHFR C677T and folate intake with ESCC risk. The heterogeneity between studies was assessed by the Q-statistics, and P values<0.1 was defined as statistical significance. Subgroup analysis was conducted to explore the possible sources of heterogeneity. If there was heterogeneity between studies, a random effect model would be used to obtain the pooled OR and its 95% CI. Otherwise, a fixed-effect model would be applied. Funnel plot an Egger's test were used to assess the publication bias. Statistical significance was defined as a two-sided P value of less than 0.05.

RESULTS
60 studies were achieved by literature search from PubMed, EMBASE and CNKI database, using different combinations of key terms. However, 39 studies were excluded from analysis, mainly due  Table-III), and a significant heterogeneity was found. However, a non-significant increased risk was found in individuals carrying CT+TT genotype. There was a significant interaction between folate intake and ESCC (P<0.05, data not shown). We also found significant heterogeneity after conducting stratification of quality of studies.
Egger's test was used to assess the publication bias, and it provided evidence that there was no publication bias among studies regarding MTHFR 677 CT and TT(P=0.46 and P=0.08, respectively).

DISCUSSION
Although previous epidemiologic studies investigated the role of MTHFR C677T for ESCC risk, the results were inconsistent. Most of these epidemiologic studies have limited sample size, and more importantly, are not power enough to give a precise estimated effect of gene variation on the ESCC risk. Therefore, meta-analysis could play a key role to obtain a more precise effect on the genetic polymorphisms on risk of disease. Three previous studies investigate the role of MTHFR C677T for ESCC risk. [29][30][31] The three meta-analyses showed individuals carrying variation of MTHFR C677T had an increased risk of ESCC, with an OR(95% CI) ranged from 1.58 to 1.78. However, the previous three meta-analyses only showed the association of MTHFR C677T on ESCC, and did not indicate the interaction between MTHFR C677T Fig.1 It is reported that folate mediates the DNA methylation reactions and function of DNA synthesis, replication, and repair, 32 which may enhance the susceptibility of cancer for individuals with low intake of folate. Previous studies have reported a high intake of food full of folate can prevent the development of various cancers. 33 Ours study has shown individuals carrying MTHFR 677TT genotype who have high intake of folate are associated with a non-significant increased risk of ESCC, and the results suggest folate intake modify the function of polymorphism in MTHFR C677T with ESCC risk.
In our study, we found significant heterogeneity between studies. However, the heterogeneity decreased or disappeared after stratifying by the intake of folate, suggesting that the folate intake was an important factor of influencing the association between variation of MTHFR C677T and ESCC risk. We found a significant heterogeneity after stratifying by quality of study, and thus this factor was not a source of heterogeneity. Some other potential factors may alter the association of this gene polymorphism and ESCC risk, such as smoking, drinking, hot food and salted food, and  they would induce between-study heterogeneity in the meta-analysis. Therefore, further studies should consider these potential risk factors. Some possible limitations should be considered in our meta-analysis. Firstly, the eligibility criteria for inclusion of ESCC patients differed for each study, which might influence the obvious consistency of effects across the included studies and cause obvious between-study heterogeneity in this metaanalysis. Secondly, there might be misclassification during our study. Some controls in our study were selected from non-cancer inpatients, and some were selected from residents. Finally, owing to lack of other genetic and environmental data of ESCC risk, the gene-environment and gene-gene interaction for esophageal cancer could not be well assessed and the outcomes from this study might be affected by risk of selection biases. Further studies are warranted to interpret their interaction.
In conclusion, our meta-analysis indicated MTHFR 677CT and/or TT genotypes are associated with the risk of ESCC, and folate could modify their association. Further large sample size and well designed study are urgently needed to further identify the association of folate intake and polymorphism of MTHFR C677T with ESCC risk.