Glutathione S-transferase M 1 Null Genotype and Hepatocellular Carcinoma Susceptibility in China and India : Evidence from an Updated Meta-analysis

Liver cancer was the most common cancer and a serious fatal disease and had caused serious damage to human health (Yu et al., 2011). The etiloogy of liver cancer was chronic infection with hepatitis B virus (HBV) and hepatitis C virus (HCV) (Jeng et al., 2014). However, some studies were about genetic polymorphisms and suggested that GSTM1 deletion polymorphism was involved in liver cancer risk in China (Wang et al., 2010; Yu et al., 2011; Chen et al., 2012; Liu et al., 2013). However, only a minority of subjects who carried GSTM1 null genotype developed liver cancer. This phenomenon suggested that other risk factors, such as other genes polymorphisms, were related to susceptibility to liver cancer (Lakkakula et al., 2013). Many studies had discussed about GSTM1 null genotype and the results were not consistent. What’s more, the researches were not performed by eliminating publication bias and irrelevant conclusions were made (Wang et al., 2010; Yu et al., 2011; Chen et al., 2012; Liu et al., 2013). We doubted that GSTM1 null genotype was the etiology of liver cancer. Glutathione-S-transferases (GSTs) were coded by glutathione S-transferase mu 1 (GSTM1), glutathione S-transferase theta-1 (GSTT1) and other genes. GSTM1 was hypothesized to protect against toxins. However, many enzymes, such as cytochrome P450 (CYP450), microsomal epoxide hydrolase, and N-acetyltransferase,


Introduction
Liver cancer was the most common cancer and a serious fatal disease and had caused serious damage to human health (Yu et al., 2011).The etiloogy of liver cancer was chronic infection with hepatitis B virus (HBV) and hepatitis C virus (HCV) (Jeng et al., 2014).However, some studies were about genetic polymorphisms and suggested that GSTM1 deletion polymorphism was involved in liver cancer risk in China (Wang et al., 2010;Yu et al., 2011;Chen et al., 2012;Liu et al., 2013).However, only a minority of subjects who carried GSTM1 null genotype developed liver cancer.This phenomenon suggested that other risk factors, such as other genes polymorphisms, were related to susceptibility to liver cancer (Lakkakula et al., 2013).Many studies had discussed about GSTM1 null genotype and the results were not consistent.What's more, the researches were not performed by eliminating publication bias and irrelevant conclusions were made (Wang et al., 2010;Yu et al., 2011;Chen et al., 2012;Liu et al., 2013).We doubted that GSTM1 null genotype was the etiology of liver cancer.

Glutathione S-transferase M1 Null Genotype and Hepatocellular
Carcinoma Susceptibility in China and India: Evidence from an Updated Meta-analysis Hong-Zhou Liu & , Jie Peng & , Chun-Yan Peng, Ming Yan, Fang Zheng* that GSTM1 null genotype was the etiology of liver cancer.Therefore, we did this meta-analysis to investigate the effect of GSTM1 null genotype in etiology of liver cancer by subgroup analysis of publication language.
We enlarged the number of cases and controls to do an undated meta-analysis, and we ruled out publication bias by using subgroup analysis of publication language to make the conclusion more convincing.

Literature inclusion criteria
(1) The subjects of literature must be Chinese and Indian; (2) The papers should include the risk of hepatocellular carcinoma and GSTM1 null genotype; (3) Only case-control and cohort studies were considered; (4) The papers must provide the sample size, the OR information such as genotype frequency that can calculate OR and 95%CI; (5) When more than one paper used the same study population, we included a recent literature.

Search strategy
PubMed and Wanfang Med Online were searched by using key words: ''liver cancer''; ''GSTM1''; ''glutathione S-transferase M1''; ''hepatocellular carcinoma''; ''polymorphism''.The date of the search interval was from 1990 to March 1, 2014 and the scope of the search was all papers consisted of journals and dissertations.

Study selection and data extraction
According to pre-established criteria of inclusion and exclusion, a double-check procedure was carried out to make sure the accuracy of the data entry.The following year, publication language, country, the data of total and exposure number in cases and control groups.A standardized procedure was performed to estimate Odds Ratio of cases and controls.Characteristics of studies were summarized.

Statistical analysis methods
Statistical analysis was did by using Review Manager5.1 and STATA 12. Adjusted OR value and 95%CI were calculated for each study, and crude OR value should be calculated if adjusted OR value was not available.The Cochrane Q statistics test and I 2 were performed for heterogeneity in this meta-analysis.A p>0.10 and I 2 <50%, simultaneously, while a random effects model was selected when p<0.10 or I 2 >50%.The funnel plot was drawn to evaluate publication bias.Egger's test and Begg's test were also done to check the publication bias.All the tests were two-sided, a P value of 0.05 for any test or model

Overview of included studies
According to the search strategy, 29 papers were selected in Figure 1.We had read all 29 the papers and 16 papers about Chinese were published by Chinese and 10 papers were published by English and 3 studies were about Indian in Table 1.

Meta-analysis Results
We observed a significant association between GSTM1 null genotype and hepatocellular carcinoma in total Chinese population [OR= 1.50, 95%CI: 1.25-1.80,p<0.0001;P Q <0.00001 and I 2 =74%] in Figure 2. As shown in Table 2, we performed subgroup analysis of publication languages to eliminate the publication bias.There was a hepatocellular carcinoma in Chinese population published by Chinese [OR= 1.74,p<0.00001;P Q =0.008 and I 2 Bonferroni correction), but not in subgroup of published by English [OR= 1.20,p=0.24;P Q <0.00001 and I 2 =84%] in Figure 4.
GSTM1 null genotype and hepatocellular carcinoma in Indian population [OR = 1.80, 95%CI: 0.80-4.02,p=0.15;P Q =0.0009 and I 2 =86%] in Figure 5. association between single GSTM1 null genotype and hepatocellular carcinoma in Indian population and Chinese population published by English.

Test of heterogeneity
Q test and I 2 were calculated to test the heterogeneity in Table 2. P value was less than 0.10, so we analyzed the pooled ORs with random effects model.Many factors might lead to heterogeneity.The distribution of GSTM1 null genotype was different in various regions; the selection of control group was different among studies; smoking and subtypes of liver cancer also could lead to heterogeneity.However, this information could not collect completely.

Publication bias
Funnel plots was performed to assess the publication bias in Figure 2, Figure 3 and Figure 4.In addition, the Egger's test and Begg's test were also selected to test publication bias in Table 3.The PBegg's test of studies published in Chinese was 0.034<0.05,so publication bias was existent.The rest of Begg's tests were p>0.05, so it indicated that there was no publication bias in other analysis.Sensitivity analysis was performed by sequential omission of individual studies, and the results also indicated that the pooled result was robust.

Discussion
with over all studies about China (OR=1.50,95%CI=1.25-1.80,p<0.0001) was observed.The homozygous deletion of GSTM1 could result in a lack of enzyme activity, so failure to deal with toxins might lead to the development of liver cancer.However, this hypothesis had a precondition that GSTM1 was the major metabolic gene or GSTM1 was the only participator in metabolize carcinogens.As we all known that there were many metabolic genes, such as cytochrome P450 (CYP450), microsomal epoxide hydrolase, and N-acetyltransferase, could metabolize carcinogens.CYP450 might have a more important role in detoxification of carcinogens, and CYP450 could compensate the non-function of GSTM1 null genotype (Liu et al., 2013).We doubt that GSTM1 null genotype was the etiology of liver cancer.
Therefore, we analyzed the results by subgroup analysis.
between GSTM1 null genotype and hepatocellular carcinoma risk in subgroup of publication in English (OR=1.20,95%CI=0.88-1.64, p=0.24).This result was because of 2141 cases and 3147 controls.However, genotype and hepatocellular carcinoma risk in subgroup of publication in Chinese (OR=1.74,95%CI=1.42-2.14, p<0.00001).Many reasons could lead to these inconsistent results.Positive results were easy to publish in Chinese journals, and this could lead to publication bias.Funnel plot, Egger's test and Begg's test were selected to test publication bias and the P Begg's test of studies published in Chinese was 0.034<0.05,so publication bias was existent.This result indicated the subgroup analysis of publication in English were more convincible than subgroup analysis of publication in Chinese.Besides, duplicate data were published among Chinese journals and publication bias was inevitable.Meta-analysis between GSTM1 null genotype and hepatocellular carcinoma risk in Indian population was performed.However, there were only 3 The heterogeneity was not negligible and it was genotype was different in various regions; the selection of control group was different among studies; smoking and subtypes of liver cancer also could lead to heterogeneity.However, this information could not collect completely.There were some limitations in this meta-analysis.First, only published papers were included in this metaanalysis, and it would cause publication bias.Second, there were a few cases and controls in Indian population to exclude and this indicates that further analysis needs to gather complete data which includes gender, age, smoking and type of liver cancer.In spite of these limitations, there were some advantages in this study.First, 3769 cases and 5517 controls included in this meta-analysis were eligible and had greater statistical power.Second, we conducted separate meta-analyses for different publication language and we derived a different and convincible conclusion of the relationship between GSTM1 and null genotype and the risk of developing liver cancer.
association between GSTM1 null genotype and the susceptibility of liver cancer by subgroup analysis of publication in English.

Figure 2 .
Figure 2. Forest Plot and Funnel Plot for the Association between GSTM1 null Genotype and Hepatocellular Carcinoma in Over all Chinese Population

Figure 4 .
Figure 4. Forest Plot and Funnel Plot for the Association between GSTM1 Null Genotype and Hepatocellular Carcinoma in Chinese Population by Publication in English