Methylenetetrahydrofolate Reductase Polymorphism C677T Is a Protective Factor for Pediatric Acute Lymphoblastic Leukemia in the Chinese Population: A Meta-Analysis

Abstract

Two polymorphisms in the methylenetetrahydrofolate reductase (MTHFR) gene, C677T and A1298C, were hypothesized to decrease the risk of acute lymphoblastic leukemia (ALL). Studies examining the associations between these two polymorphisms and ALL susceptibility drew inconsistent results. To obtain a reliable conclusion in a Chinese population, we carried out a meta-analysis. In total, 11 studies on C677T polymorphism (1597 cases and 2295 controls) and 10 studies on A1298C polymorphism (1553 cases and 2224 controls) were included in the meta-analysis. We found a significant association between the 677T variant and reduced ALL risk in Chinese children (Dominant model: odds ratio [OR_FE]=0.73, 95% confidence interval [CI]: 0.63-0.86, p<0.01). Heterogeneity between the studies in the children subgroup was weak and vanished after excluding one study deviating from HWE in the control group (p>0.1). In the adult subgroup, there was no significant association between the C677T variant and ALL risk (Dominant model: OR_RE=0.88, 95% CI: 0.45-1.72, p=0.72). Significant heterogeneity was found in the adult subgroup in all the genetic model tests (p<0.1). The A1298C polymorphism had an effect on ALL risk neither in adults (Dominant model: OR_FE=0.95, 95% CI: 0.71-1.27, p=0.72) nor in children (Dominant model: OR_FE=1.02, 95% CI: 0.87-1.21, p=0.77). No significant heterogeneity between studies on A1298C polymorphism was found in the meta-analysis (p>0.1). The results showed that there was a protective effect of the MTHFR C677T variant on ALL risk in Chinese children.

Introduction

Acute lymphoblastic leukemia (ALL) is a common cancer, especially in children. The disease-free survival of childhood ALL has exceeded 80% in the developed countries over the last years. However, almost 20% of the children with ALL either relapse or do not respond to treatment (Karathanasis et al., 2011). A variety of factors might be related to the biologic mechanisms and etiology of ALL. It is generally considered that the development of ALL is a comprehensive result of environment, genetic risk factors, and gene-environment interactions (Robien and Ulrich 2003; Scelo et al., 2009). Folate deficiency and aberrant metabolism have been reported to be associated with ALL (Yang et al., 2011). Polymorphisms in genes involved in folate transport, metabolism, and distribution in vivo drew widespread attention in the last decade (McNeer, 2011). Two polymorphisms in the methylenetetrahydrofolate reductase (MTHFR) gene, C677T and A1298C, were hypothesized to decrease the risk of ALL because they reduced MTHFR enzyme activity, leading to enhanced availability of 5,10-methylenetetrahydrofolate in the DNA synthesis pathway and reduced uracil misincorporation into DNA (Skibola et al., 1999).

A large literature examined the associations between C677T and A1298C and ALL susceptibility. However, there were inconsistent results. Several meta-analyses were carried out to resolve these discrepancies in the last few years (Wang et al., 2010b; Tong et al., 2011; Yan et al., 2012; Zintzaras et al., 2012). For the A1298C variant, there were mainly negative results found in these meta-analyses. For the C677T variant, these meta-analyses yield relatively consistent results in Caucasians (Whites), but not in Asians. Among them, Yan et al. (2012) reported a protective effect of the 677T polymorphism in Asian children. It was inconsistent with the study by Zintzaras et al. (2012), which claimed no significant effect of C677T variant on ALL risk in East Asians. A meta-analysis provided data on the whole population of Asians, including both children and adults, also denied a significant role of C677T polymorphism in Asians (Tong et al., 2011).

The discrepancy in conclusions might be caused by different grouping criteria. To our knowledge, ethnicity and lifestyle are different among populations in Asian countries. The Asian population (Tong et al., 2011), thereby, is not a strict concept in the genetic study (Zintzaras et al., 2012). It might be more reasonable to include only Chinese and Koreans in the East Asian subgroup (Wang et al., 2010b). At the same time, language bias might also contribute to this discrepancy, as there were different databases and language limitations used in these studies. Furthermore, using different genetic models in these meta-analyses was also one of the reasons for the conflicting results in Asians.

As there were plenty of studies on the Chinese population published in recent years, we carried out a meta-analysis to provide more information on the subject. We endeavored to include all the studies on MTHFR C677T, A1298C, and ALL susceptibility in the Chinese population, without any restriction on language, to decrease language bias and publication bias as much as possible.

Methods

Selection criteria and identification of studies

PUBMED, EMBASE, China Science and Technology journal database, Chinese Biomedicine database (CBM), China National Knowledge Infrastructure platform (CNKI), and Wanfang database were searched using the following search terms: methylenetetrahydrofolate reductase or MTHFR; ALL; polymorphism or genetic variant; and China or Chinese, without any restriction on language. The latest searches were undertaken on Feb 28, 2012. All the references cited in the eligible studies were reviewed to identify additional publications.

All the case-control studies providing valid data on the association between the MTHFR polymorphism (C677T or A1298C) and ALL risk in Chinese were included in the meta-analysis. When multiple articles reported on the same population, we used the most recent one or the one with the largest number of participants only.

Data collection

Two authors extracted the following data from the included studies independently: first author's name, year of publication, participant demographics (age and sex), genotype distribution, leukemia characteristics. Discrepancies were resolved by discussion, when necessary, adjudicated by a third reviewer.

Statistical analysis

The meta-analysis examined the overall association of the MTHFR C677T or A1298C polymorphisms with the risk of ALL in Chinese by odds ratio (OR) with corresponding 95% confidence intervals (CIs) under allele contrast, dominant model, recessive model, additive model, and heterozygote versus wild-type homozygous (CT vs. CC or AC vs. AA) (Zintzaras and Lau, 2008; Tong et al., 2011). Stratification analysis was carried out by age (adults or children).The presence of heterogeneity between studies was assessed using Cochran's Q statistic (Zintzaras and Ioannidis, 2005). p<0.10 indicated significant heterogeneity between studies; then the degree of heterogeneity was quantified with I² metric [I²=(Q-d.f.)/Q], I²>30% suggesting evidence for heterogeneity (Higgins and Thompson, 2002). If the studies were not statistically heterogeneous, the pooled OR was estimated using the fixed-effects model (Mantel-Haenszel). Alternatively, the pooled OR was estimated using the random-effects model (DerSimonian Laird)(Zintzaras et al., 2006). Potential publication bias was evaluated using the Egger's regression test and the Begg-Mazumdar test (p<0.10 indicating statistically significant)(Macaskill et al., 2001). Hardy-Weinberg equilibrium (HWE) in the control group for all studies was evaluated using the Pearson's χ² test. Sensitivity analysis was carried out by excluding specific studies with controls not in HWE (Trikalinos et al., 2006). The influence of individual study was evaluated by estimating the pooled ORs in the absence of each study (Wang et al., 2010a). All the p-values were two-sided, p<0.05 indicating significant unless otherwise stated, and all analyses were performed with Stata software (version 10.0; Stata Corporation, College Station, TX).

Results

Eligible studies

Ten eligible published articles providing data on the relationship between MTHFR C677T, A1298C, and ALL in Chinese population were included in the meta-analysis. The studies were published between 2004 and 2011. Nine articles dealt with both C677T and A1298C, and one dealt with C677T only (Yu, 2010). One article (Yang et al., 2011) provided data on adults and children, respectively. There were 11 studies on C677T and 10 studies on A1298C, accordingly. Among them, three studies (Zhang, 2007; Lv et al., 2010; Yang et al., 2011) dealt with adult ALL patients, and seven dealt with pediatric patients (Jiang et al., 2004; Yu, 2004; Tong et al., 2010; Yeoh et al., 2010; Yu, 2010; Yang et al., 2011; Lv et al., 2011), one study included a mixed population of adult and pediatric patients (Liu et al., 2008). All the included studies, except for that by Jiang et al. (2004), described the diagnostic criteria, according to morphologic and immunophenotypic analysis. Controls were age and gender matched with ALL patients in all the included studies. No study-mentioned genotyping was performed under blind conditions. The control groups were either hospital outpatients without history of malignant diseases or community healthy controls. Genotype distributions among the controls of all studies were in agreement with HWE, except that by Jiang et al. (2004). Detailed information of the included studies is available in Table 1.

Table 1.

Characteristics of Eligible Studies Included in the Meta-Analysis

						C677T		A1298C
Author	Year	Study population	Source of controls	No. of cases (M/F)	No. of controls (M/F)	MAF	HWE	MAF	HWE	Subtypes of ALL
Yang L(A)	2011	Adults	Population	130(78/52)	367(210/157)	0.54	0.14	0.16	0.28	Not described
Yang L(C)	2011	Children	Population	231(140/91)	367(210/157)	0.54	0.14	0.16	0.28	Not described
Lv H	2011	Children	Hospital	176(108/68)	170(101/69)	0.46	0.68	0.15	0.27	Not described
Yu J	2010	Children	Hospital	45(28/17)	70(45/25)	0.35	0.82	-	-	Not described
Tong N	2010	Children	Population	316(216/145)	508(318/190)	0.41	0.27	0.19	0.07	84.2% B-ALL and 15.8% T-ALL
Yeoh AE	2010	Children	Population	321(185/136)	346(156/190)	0.31	0.77	0.22	0.65	89.4% B-ALL,7.8% T-ALL and 2.8% infant
Lv L	2010	Adults	Hospital	127(59/68)	182(85/97)	0.37	0.7	0.15	0.27	100% B-ALL
Liu JX	2008	Adults+Children	Population	83(60/23)	83(54/29)	0.33	0.91	0.17	0.68	Not described
Zhang J	2007	Adults	Population	46(27/19)	80(53/27)	0.42	0.64	0.21	0.68	Not described
Yu H	2004	Children	Population	51(37/14)	53	0.41	0.47	0.31	0.58	72.5% B-ALL and 27.5% non B-ALL
Jiang H	2004	Children	Population	30(18/12)	70(38/32)	0.43	0.02	0.19	0.74	Not described

A, data on adult population; C, data on children population; HWE, Hardy-Weinberg equilibrium; MAF, minor allele frequency among controls; ALL, acute lymphoblastic leukemia.

C677T polymorphism and ALL risk

In total, there were 11 studies on the C677T polymorphism included in the meta-analysis, containing 1597 ALL cases and 2295 controls. The mean T allele frequency among Chinese was 0.42. The pooled results showed that the protective role of 677T in Chinese population was weak (Dominant model: OR_RE=0.77, 95% CI: 0.61-0.97, p=0.03; CT vs. CC: OR_RE=0.76, 95% CI: 0.59-0.96, p=0.03). The marginally significant association disappeared in the sensitivity analysis after exclusion of one study deviating from HWE in the control group (Dominant model: OR_RE=0.80, 95% CI: 0.64-1.01, p=0.056; CT vs. CC: OR_RE=0.79, 95% CI: 0.62-1.00, p=0.053). Significant heterogeneity existed between the studies (Table 2 and Fig. 1). However, significant results were observed in the children subgroup, with the protective effect of 677T polymorphism found in all the genetic model tests except in the recessive model test (Dominant model: OR_FE=0.73, 95% CI: 0.63-0.86, p<0.01; Recessive model: OR_FE=0.86, 95% CI: 0.70-1.06, p=0.16). Exclusion of one study with controls deviating from HWE in the sensitivity study brought about no significant change in the children's subgroup. Moreover, heterogeneity between the studies in the children group was weak (p=0.065 in allele contrast and p=0.097 in CT vs. CC comparison), and vanished after excluding one study (Jiang et al., 2004) deviating from HWE in the control group (p>0.1 in all the genetic model tests). In subgroup analysis on adults, no significant association was observed in any genetic model test. Significant heterogeneity was found in the adult subgroup in all the genetic model tests. The results of the individual study influence suggested that there was no single study exhibiting excessive study influence. Significant publication bias was detected neither in the Egger's regression test (p=0.237) nor in the Begg-Mazumdar test (p=0.213).

FIG. 1.

Odds ratio (OR) estimates with the corresponding 95% confidence intervals [CIS] for the dominant model test of MTHFR C677T polymorphism and the risk of ALL in Chinese. A, data on adults; C, data on children.

Table 2.

Odds Ratios and Heterogeneity Results for the Genetic Contrasts of MTHFR C677T Polymorphism for All Risk in Chinese Population

C677T contrast		Studies	I², (%)	p, Q Test^a	Fixed-effect OR	Random-effect OR
Dominant model	All	11	58.3	0.008	0.80(0.70-0.92)	0.77(0.61-0.97)
	All in HWE	10	55.0	0.018	0.82(0.71-0.94)	0.80(0.64-1.01)
	Adults	3	77.1	0.013	1.00(0.74-1.35)	0.88(0.45-1.72)
	Children	7	42.3	0.109	0.73(0.63-0.86)	0.70(0.55-0.88)
	Children in HWE	6	29.8	0.212	0.75(0.64-0.89)	0.74(0.60-0.91)
Recessive model	All	11	69.8	0.000	0.86(0.72-1.02)	0.88(0.61-1.27)
	All in HWE	10	71.3	0.000	0.87(0.73-1.04)	0.91(0.63-1.31)
	Adults	3	78.0	0.011	0.63(0.44-0.90)	0.66(0.27-1.58)
	Children	7	39.6	0.127	0.86(0.70-1.06)	0.85(0.63-1.16)
	Children in HWE	6	38.6	0.149	0.88(0.71-1.09)	0.87(0.65-1.17)
Allele contrast	All	11	69.9	0.000	0.86(0.78-0.95)	0.84(0.69-1.01)
	All in HWE	10	69.0	0.001	0.87(0.79-0.96)	0.87(0.72-1.04)
	Adults	3	82.9	0.003	0.86(0.71-1.05)	0.81(0.48-1.36)
	Children	7	49.4	0.065	0.82(0.74-0.92)	0.79(0.67-0.94)
	Children in HWE	6	38.7	0.148	0.82(0.74-0.92)	0.83(0.71-0.97)
Additive model	All	11	63.7	0.002	0.80(0.65-0.97)	0.79(0.54-1.14)
	All in HWE	10	63.8	0.003	0.82(0.67-1.00)	0.82(0.57-1.18)
	Adults	3	78.7	0.009	0.75(0.49-1.14)	0.68(0.25-1.85)
	Children	7	27.1	0.222	0.72(0.57-0.92)	0.72(0.53-0.98)
	Children in HWE	6	12.2	0.337	0.75(0.59-0.95)	0.75(0.57-0.98)
CT vs. CC	All	11	56.5	0.011	0.81(0.70-0.94)	0.76(0.59-0.96)
	All in HWE	10	55.7	0.016	0.82(0.71-0.96)	0.79(0.62-1.00)
	Adults	3	71.7	0.029	1.13(0.82-1.55)	1.00(0.53-1.88)
	Children	7	44.1	0.097	0.74(0.63-0.88)	0.69(0.54-0.89)
	Children in HWE	6	41.9	0.126	0.76(0.64-0.90)	0.72(0.57-0.93)

All, all of the studies meet the inclusion criteria.

Q test is to estimate heterogeneity between studies, p<0.10 indicates significant heterogeneity.

OR, odds ratio.

A1298C polymorphism and ALL risk

Considering the A1298C variant, there were 10 studies on this polymorphism included in the meta-analysis, containing 1553 ALL cases and 2224 controls. The mean C allele frequency among the Chinese was 0.19. No significant association between the A1298C variant and ALL susceptibility was found in any genetic model test in the Chinese population (Dominant model: OR_FE=0.98, 95% CI: 0.86-1.13, p=0.81). Similar results were obtained in adults (Dominant model: OR_FE=0.95, 95% CI: 0.71-1.27, p=0.72) and children (Dominant model: OR_FE=1.02, 95% CI: 0.87-1.21, p=0.77) subgroup analysis. More importantly, heterogeneity among studies on the Chinese population was not significant in all genetic contrasts in all the study subgroups (Table 3 and Fig. 2). The removal of any one study did not result in a movement of the point estimate outside the 95% CIs. The results suggested that there was no single study, which exhibited an excessive study influence. There was no significant publication bias detected by formal statistics (Begg's test, p=0.93; Egger's test, p=0.85).

FIG. 2.

OR estimates with the corresponding 95% CIS for the dominant model test of MTHFR A1298C polymorphism and the risk of ALL in Chinese. A, data on adults; C, data on children.

Table 3.

Odds Radios and Heterogeneity Results for the Genetic Contrasts of A1298C Polymorphism for All Risk in Chinese Population

A1298C contrast		Studies	I², (%)	p, Q Test^a	Fixed-effect OR	Random-effect OR
Dominant model	All	10	32.2	0.151	0.98(0.86-1.13)	0.98(0.82-1.17)
	Adults	3	7.4	0.340	0.95(0.71-1.27)	0.95(0.70-1.29)
	Children	6	41.2	0.130	1.02(0.87-1.21)	1.04(0.83-1.31)
Recessive model	All	10	0	0.805	1.10(0.78-1.55)	1.11(0.78-1.58)
	Adults	3	0	0.511	1.33(0.59-3.03)	1.40(0.60-3.26)
	Children	6	0	0.605	1.06(0.72-1.56)	1.06(0.71-1.57)
Allele contrast	All	10	29.5	0.173	1.00(0.89-1.12)	0.99(0.86-1.15)
	Adults	3	0	0.402	0.99(0.77-1.27)	0.99(0.77-1.27)
	Children	6	43.7	0.114	1.02(0.89-1.18)	1.03(0.84-1.26)
Additive model	All	10	0	0.718	1.10(0.78-1.57)	1.12(0.78-1.60)
	Adults	3	0	0.521	1.29(0.56-2.95)	1.36(0.58-3.18)
	Children	6	0	0.463	1.08(0.73-1.60)	1.08(0.73-1.62)
CA vs. AA	All	10	25.9	0.205	0.97(0.84-1.12)	0.97(0.81-1.18)
	Adults	3	13.5	0.315	0.92(0.68-1.25)	0.92(0.66-1.28)
	Children	6	28.9	0.218	1.02(0.86-1.21)	1.03(0.83-1.28)

All=all of the studies meet the inclusion criteria.

Q test is to estimate heterogeneity between studies, p<0.10 indicates significant heterogeneity.

Discussion

Our meta-analysis summarized all the available data on the association between MTHFR C677T and A1298C polymorphism and ALL risk in the Chinese population. The results showed a protective effect of C677T variant on ALL risk in Chinese children. There might be a similar effect of the C677T polymorphism on ALL risk in children population with different genetic background, especially in Caucasians and Chinese. Environmental factors played an import role in the C677T polymorphism effect on ALL risk. The protective effect of MTHFR polymorphisms against ALL might be influenced by folate state in vivo. When the folate intake is deficient, both DNA methylation and thymidylate synthesis may be impaired in patients with MTHFR polymorphisms, leading to carcinogenesis (Sadananda et al., 2010). That might be one of the reasons for no significant association found in the ethnicity other than Caucasians and Asians (Yan et al., 2012).

Considering the effect of MTHFR C677T polymorphism on adult ALL risk, our result was different from previous meta-analyses. Pereira et al. (2006) found that the C677T polymorphism was related to a decreased risk of adult ALL (Recessive model: OR=0.45, 95% CI: 0.26-0.77; n=3). Tong et al. (2011) reported a significantly reduced risk of adult ALL in a dominant model comparison (OR=0.89, 95%CI: 0.80-0.98; n=4). There was no association between the variant and ALL risk in adults observed in any genetic model test in our study. These results of meta-analysis in the adult subgroup were not stable owing to the limited number of studies included and significant heterogeneity among these studies. Therefore, the existing evidence was insufficient to confirm or deny the role of C677T variant in adults. A series of differences existed between children's and adult's ALL. Compared to adults, children were in a period of rapid cell division and growth and had a higher folate requirement. At the same time, they were not exposed to many environmentally carcinogenic factors, and might be influenced more by diet- and genetic-related etiology factors (Koppen et al., 2010). These differences might contribute to the different results between children and adults. It was easy to understand the protective effect of 677T was more obvious in children in our study. Of course, there were only a few studies on adult ALL, a more confirmed meta-analysis result requires more studies on this age group.

There was no significant association between the A1298C variant and ALL susceptibility in the Chinese population found in our study. Reviewing the meta-analyses published so far, with the increasing number of studies included, there were mainly insignificant results reported on A1298C, except by Tong et al. (2011) who reported an increased ALL risk in children in the comparison of 1298CA versus the AA genotype (n=16). It was noteworthy that the significant effect of A1298C polymorphism was not a stable result. We have pooled all the studies in English language journals evaluating A1298C and ALL risk published from 1999 through 2011 (n=29) and found no significant result in any age subgroup (children ALL n=22; adult ALL n=5) or ethnicity subgroup (Caucasian n=12; East Asian n=7) (data not shown). Therefore, comprehensively analyzing these results, we tended to consider there was no significant effect of the A1298C polymorphism on ALL risk. Compared to the C677T polymorphism, the A1298C polymorphism contributed less effect on enzyme activity reduction (Friedman et al., 1999; Weisberg et al., 2001), which might not be powerful enough to affect ALL susceptibility.

Our meta-analysis carried out on studies in a single ethnicity and provided more information on the association between MTHFR polymorphisms and ALL susceptibility. The results indicated that there was a protective effect of theC677T polymorphism on children's ALL risk and this protective role was similar in different ethnicities, especially between the Caucasians and Chinese. The pooled results did not support a significant effect of the C677T polymorphism on adult ALL risk. However, the reliability of results on adults in our meta-analysis, as well as the previous meta-analysis, was limited by the few studies included and significant heterogeneity existing between these studies. Thus, more large and well-designed studies on adult ALL were needed. Moreover, the association between MTHFR polymorphisms and ALL risk was influenced by the gene-gene and gene-environment interactions. Most of the studies included in our meta-analysis did not provide the relevant information. We therefore appeal for further studies providing more information on gene-gene and gene-environment interactions.

Footnotes

Author Disclosure Statement

None declared.

References

Friedman

, Goldschmidt

, Friedlander

et al. 1999. A common mutation A1298C in human methylenetetrahydrofolate reductase gene: association with plasma total homocysteine and folate concentrations. J Nutr, 129:1656-1661.

Higgins

, Thompson

. 2002. Quantifying heterogeneity in a meta-analysis. Stat Med, 21:1539-1558.

Jiang

, Xue

, Pan

et al. 2004. Methylenetetrahydrofolate reductase gene polymorphism in childhood acute lymphocytic leukaemia. Zhonghua Xue Ye Xue Za Zhi, 7:57-58.

Karathanasis

, Stiakaki

, Goulielmos

et al. 2011. The role of the methylenetetrahydrofolate reductase 677 and 1298 polymorphisms in Cretan children with acute lymphoblastic leukemia. Genet Test Mol Biomarkers, 15:5-10.

Koppen

, Hermans

, Kaspers

. 2010. Folate related gene polymorphisms and susceptibility to develop childhood acute lymphoblastic leukaemia. Br J Haematol, 148:3-14.

Liu

, Chen

, Lin

. 2008. Relationship between genetic polymorphism of methylenetetrahydrofolate reductase and the risk of acute lymphocytic leukemia. Med J Chin PLA, 11:1291-1293.

, Wang

, Du

et al. 2011. Relationship between genetic polymorphism of methylenetetrah∼ drofolate reduetase and the risk of childhood acute lymphocytic leukemia. J Clin Pediatr, 5:414-417.

, Wu

, Sun

et al. 2010. Combined 677CC/1298AC genotypes of methylenetetrahydrofolate reductase (MTHFR) reduce susceptibility to precursor B lymphoblastic leukemia in a Chinese population. Eur J Haematol, 84:506-512.

Macaskill

, Walter

, Irwig

. 2001. A comparison of methods to detect publication bias in meta-analysis. Stat Med, 20:641-654.

10.

McNeer

. 2011. The complex interplay between folate metabolism and risk of acute lymphoblastic leukemia. Leuk Lymphoma, 52:1621-1622.

11.

Pereira

, Rudnicki

, Pereira

et al. 2006. 5,10-Methylenetetrahydrofolate reductase polymorphisms and acute lymphoblasticleukemia risk: a meta-analysis. Cancer Epidemiol Biomarkers Prev, 15:1956-1963.

12.

Robien

, Ulrich

. 2003. 5,10-Methylenetetrahydrofolate reductase polymorphisms and leukemia risk: a HuGE minireview. Am J Epidemiol, 157:571-582.

13.

Sadananda

AMN

, Chandy

, Ramachandra

et al. 2010. Methylenetetrahydrofolate reductase gene polymorphisms and risk of acute lymphoblastic leukemia in children. Indian J Cancer, 47:40-45.

14.

Scelo

, Metayer

, Zhang

et al. 2009. Household exposure to paint and petroleum solvents, chromosomal translocations, and the risk of childhood leukemia. Environ Health Perspect, 117:133-139.

15.

Skibola

, Smith

, Kane

et al. 1999. Polymorphisms in the methylenetetrahydrofolate reductase gene are associated with susceptibility to acute leukemia in adults. Proc Natl Acad Sci U S A, 96:12810-12815.

16.

Tong

, Fang

, Li

et al. 2010. Methylenetetrahydrofolate reductase polymorphisms, serum methylenetetrahydrofolate reductase levels, and risk of childhood acute lymphoblastic leukemia in a Chinese population. Cancer Sci, 101:782-786.

17.

Tong

, Sheng

, Wang

et al. 2011. Methylenetetrahydrofolate reductase gene polymorphisms and acute lymphoblasticleukemia risk: a meta-analysis based on 28 case-control studies. Leuk Lymphoma, 52:1949-1960.

18.

Trikalinos

, Salanti

, Khoury

et al. 2006. Impact of violations and deviations in Hardy-Weinberg equilibrium on postulatedgene-disease associations. Am J Epidemiol, 163:300-309.

19.

Wang

, Wang

, Chang

et al. 2010a. Endothelial nitric oxide synthase gene polymorphisms and essential hypertension in Han Chinese. Genet Mol Res, 9:1896-1907.

20.

Wang

, Zhan

, Chen

et al. 2010b. MTHFR C677T polymorphisms and childhood acute lymphoblastic leukemia: ameta-analysis. Leuk Res, 34:1596-1600.

21.

Weisberg

, Jacques

, Selhub

et al. 2001. The 1298A—>C polymorphism in methylenetetrahydrofolate reductase (MTHFR): in vitro expression and association with homocysteine. Atherosclerosis, 156:409-415.

22.

Yan

, Yin

, Dreyer

et al. 2012. A meta-analysis of MTHFR C677T and A1298C polymorphisms and risk of acute lymphoblastic leukemia in children. Pediatr Blood Cancer, 58:513-518.

23.

Yang

, Liu

, Wang

et al. 2011. Polymorphisms in folate-related genes: impact on risk of adult acute lymphoblastic leukemia rather than pediatric in Han Chinese. Leuk Lymphoma., 52:1770-1706.

24.

Yeoh

, Lu

, Chan

et al. 2010. Genetic susceptibility to childhood acute lymphoblastic leukemia shows protection in Malay boys: results from the Malaysia-Singapore ALL Study Group. Leuk Res, 34:276-283.

25.

. 2004. The Relationship Between the MTHFR Gene Polymorphisms and Childhood Acute Lymphocytic Leukemia as Well as the Toxicity of MTX [D] Huazhong University of Science and Technology: Wuhan.

26.

. 2010. Association of Drug Metabolizing Enzymes Genetic Polymorphisms with the Susceptibility and Methotrexateinduced Toxicities in Childhood Acute Leukemia [D] Southern Medical University: Guangdong.

27.

Zhang

. 2007. A Study on the Serum Folate Level and Methylenetetrahydrofolate Reductase Genetic Polymorphisms in Adult Acute Lymphocytic Leukemia[D] Hebei Medical University: Shijiazhuang.

28.

Zintzaras

, Doxani

, Rodopoulou

et al. 2012. Variants of the MTHFR gene and susceptibility to acute lymphoblastic leukemia in children: a synthesis of genetic association studies. Cancer Epidemiol, 36:169-176.

29.

Zintzaras

, Ioannidis

. 2005. Heterogeneity testing in meta-analysis of genome searches. Genet Epidemiol, 28:123-137.

30.

Zintzaras

, Kitsios

, Stefanidis

. 2006. Endothelial NO synthase gene polymorphisms and hypertension: a meta-analysis. Hypertension, 48:700-710.

31.

Zintzaras

, Lau

. 2008. Synthesis of genetic association studies for pertinent gene-disease associations requires appropriate methodological and statistical approaches. J Clin Epidemiol., 61:634-645.