Leukemia & Lymphoma, October 2006; 47(10): 2070 – 2075
ORIGINAL ARTICLE: CLINICAL
Association between the MTHFR A1298C polymorphism and increased risk of acute myeloid leukemia in Brazilian children
FLAVIO JOSE DA COSTA RAMOS1,2, MARIA TEREZA CARTAXO MUNIZ2, ´ JO3, EDNALVA PEREIRA LEITE4, VANESSA CAVALCANTE SILVA1, MARCELA ARAU 2 ELIZABETE MALAQUIAS FREITAS , CRISIANE WAIS ZANROSSO5, ANA HATAGIMA6, MARICILDA PALANDI DE MELLO7, JOSE ANDRE´S YUNES3, TEREZINHA DE JESUS MARQUES-SALLES4, NEIDE SANTOS1, SILVIA R. BRANDALISE3, & MARIA S. POMBO-DE-OLIVEIRA5 1
Departamento de Gene´tica, Universidade Federal de Pernambuco, Recife, Pernambuco, Brazil, 2Departmento de Biologia e Departamento de Cieˆncias Fisiolo´gicas, Instituto de Cieˆncias Biologicas, Universidade de Pernambuco, Recife, Pernambuco, Brazil, 3Centro Infantil Boldrini, Campinas, Sa˜o Paulo, Brazil, 4Centro de Oncologia, Hospital Universita´rio Oswaldo Cruz, Recife, Pernambuco, Brazil, 5CPq-Instituto Nacional de Caˆncer, Divisa˜o de Medicina Experimental, Rio de Janeiro, Brazil, 6 Departamento de Gene´tica, Instituto Oswaldo Cruz, Fundac¸a˜o Oswaldo Cruz, Rio de Janeiro, Brazil, and 7Centro de Biologia Molecular e Engenharia Gene´tica, Universidade Estadual de Campinas, Sa˜o Paulo, Brazil (Received 25 March 2006; accepted 11 May 2006)
Abstract Methylenetetrahydrofolate reductase (MTHFR) is an essential enzyme in the metabolism of folate. The presence of polymorphisms that reduce the activity of MTHFR has been linked to the multifactor process of development of acute leukemia. A case control study was conducted on Brazilian children in different regions of the country with the aim of investigating the role of MTHFR C677T and A1298C polymorphisms as risk factors in the development of acute myeloid leukemia (AML). We used the polymerase chain reaction restriction fragment length polymorphism method to genotyping 182 AML and 315 healthy individuals. The genotype 677 CT was associated with decreased risk [odds ratio (OR), 0.37; confidence interval (CI) 95%, 0.14 – 0.92], whereas 1298 AC genotype was linked with an increased risk [OR, 2.90; CI 95%, 1.26 – 6.71] of developing AML in non-white children. Further epidemiological study is needed to unravel the complex multiple gene-environment interactions in the role of the AML leukemogenesis.
Keywords: Acute myeloid leukemia, MTHFR, polymorphism, childhood, AML, Brazil
Introduction Acute myeloid leukemia (AML) is characterized by an increase in the proliferation of myeloid cells and their alteration in the different maturation steps. Currently, AML corresponds to approximately 15 – 17% of cases of childhood leukemia. The AML subtypes are often characterized by genetic alterations such as scattered mutations and chromosomal alterations, including translocations, inversions, and deletions [1]. There have been a number of studies aiming to identify the genetic and environmental risk
factors associated with the leukemogenesis processes of AMLs, as well as the prognostic factors involved in this disease [2]. Folic acid is an important vitamin for cell division and for maintaining homeostasis. A deficiency of folic acid results in a number of cellular disorders [3]. Byproducts of its metabolism take part in processes of methylation and also participate in the synthesis and repair of DNA, a mechanism that is vital in the prevention of chromosomal abnormalities [4]. The enzyme methylenetetrahydrofolate reductase (MTHFR) plays an essential role in the metabolism
Correspondence: Maria S. Pombo-de-Oliveira, Instituto Nacional de Caˆncer Divisa˜o de Medicina Experimental Rua Andre´ Cavalcanti, 37 CEP: 20231 050, Rio de Janeiro, Brazil. Tel: 0055 21 3233 1324. Fax: 0055 21 3233 1470. E-mail:
[email protected] ISSN 1042-8194 print/ISSN 1029-2403 online Ó 2006 Informa UK Ltd. DOI: 10.1080/10428190600800132
MTHFR polymorphisms in childhood AML of folic acid, irreversibly converting 5,10-methylenetetrahydrofolate into 5-methyltetrahydrofolate, a condition required for the conversion of homocysteine to methionine [5]. In the last few years, studies have shown an association between two polymorphisms in the gene MTHFR (C677T and A1298C) with the incidence of cases of acute lymphoid leukemia (ALL) in children and adults [6 – 10]. These two polymorphisms, C677T and A1298C, have been associated with reduced enzyme activity of MTHFR which consequently leads to an accumulation of 5,10methylenetetrahydrofolate, which in turn will be relocated towards synthesizing purines, converting dUMP to dTMP and repairing DNA damages [11]. These reactions are associated to the prevention of genetic alterations that trigger leukemogenesis. With this reasoning, the aim of this study was to investigate the association between the MTHFR polymorphisms C677T and A1298C and the risk of developing AML in a group of children from different regions in Brazil.
Design and methods Subjects Samples from a series of 182 children and teenagers with AML were subjected to genetic polymorphism analysis in a case-control study. They were treated at the following medical centers in Brazil: i) Centro de Oncologia do Hospital Universita´rio Oswaldo Cruz, Recife, Pernambuco (located in the northeast region); ii) Instituto Nacional de Caˆncer, Rio de Janeiro; and iii) Centro Infantil Boldrini, Campinas, Sa˜o Paulo (both located in the southeast region). Apart from these three main reference centers, minority cases of AML in the south and center-west regions were also included in this study. Criteria The diagnosis of AML was based on the FrenchAmerican-British group criteria [12]. Cases of AML secondary to previous treatment and/or with associated myelodysplasia were excluded from the study.
2071
history of cancer who lived in the same regions as the children with AML participating in this study. The ethics committee of the institutions involved approved the study. Methods Genomic DNA was isolated from peripheral white blood cells using a modified protocol previously described by Miller et al. [14]. The C677T polymorphism was determined by a polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) assay [15]. The C!T transition creates a restriction site for the enzyme HinfI and the digested product was isolated electrophoretically in 2% agarose gel and the fragments were visualized in ultraviolet light (UV) after being stained with ethidium bromide. Wild type (CC) shows a single fragment of 198 bp; heterozygote (CT) shows fragments of 198, 175 and 23 bp; and mutant homozygote (TT) shows two fragments with 175 and 23 bp. To determine the MTHFR A1298C polymorphism a modified PCR-RFLP protocol was used as previously described by Skibola et al. [16]. The A!C transversion eliminates a restriction site for the enzyme MboII. Visualization was conducted after staining the gel with ethidium bromide and exposing it to UV light. Wild types (AA) show five fragments of 56, 31, 30, 28 and 18 bp; heterozygotes (AC) show six fragments of 84, 56, 31, 30, 28, and 18 bp; mutant homozygotes (CC) show four fragments of 84, 31, 30 and 18 bp. Statistical analysis The maximum likelihood method was used to estimate allelic frequencies. Goodness-of-fit of genotype distribution to Hardy – Weinberg equilibrium was ascertained by the chi-square test. Comparison among race, AML subtypes and other variables was performed through contingency tables analyzed by chi-square value. Analyses were performed matching cases and controls according to racial and geographic regions. The analysis was carried out using the statistical package SPSS (version 13.0) to estimate odds ratio (OR) and confidence interval at 95% (CI 95%) significance level.
Race Because of the complexity to define the heterogeneity of the Brazilians’ ethnicity, participants of this study was categorized only in two groups: white and nonwhite according to criteria described previously [7,13]. The control group consisted of peripheral blood samples obtained from 315 subjects with no previous
Results The main characteristics of case and control samples are presented in Table I. The majority of the children in this series, either cases or controls, came from urban or semi-urban environments surrounding the largest participating cities, with population sizes usually over 1 million people. There were a
2072
F. J. Da Costa Ramos et al.
Table I. Demographic characteristics of the AML series and the allele frequency of the MTHFR gene. Characteristics Sex Male Female Race White Non-whites AML subtypes M0-M1 M2 M3 M4-M5 M6-M7 NK
AML n ¼ 182 (%)
Controls n ¼ 315 (%)
89 (49.0) 93 (51.0)
151 (48.0) 164 (52.0)
139 (76.0) 43 (24.0)
209 (66.0) 106 (34.0)
19 43 30 47 30 13
(10.4) (23.6) (16.4) (25.8) (16.4) (7.1)
Table II. Hardy-Weinberg equilibrium analysis and allele frequency of MTHFR C677T and A1298C polymorphism. n
w21
315 182 43 30 47 49
A1298C Controls AMLs M2 M3 M4-M5 Others*
C677T Controls AMLs M2 M3 M4-M5 Others*
P
C allele
T allele
0.395 3.151 0.105 0.578 2.754 3.639
0.056**
0.698 0.695 0.714 0.633 0.685 0.696
0.302 0.305 0.286 0.367 0.315 0.304
n
w21
P
A allele
C allele
315 182 43 30 47 49
1.669 2.234 0.436 0.140 1.746 4.667
0.031
0.768 0.742 0.726 0.767 0.761 0.707
0.232 0.258 0.274 0.233 0.239 0.293
NA
Regions Southeast Northeast Others*
131 (71.9) 34 (18.7) 17 (9.3)
150 (47.6) 93 (29.5) 72 (23.0)
Family Income** 400 – 999.00 1,000 – 1,999.00 2,000 or more
129 (70.8) 35 (19.2) 18 (9.9)
223 (70.2) 50 (15.8) 42 (13.3)
C667T CC CT TT
93 (51.0) 67 (37.0) 22 (12.0)
156 (49.5) 128 (40.5) 31 (10.0)
A1298C AA AC CC
104 (57.1) 62 (34.1) 16 (8.8)
190 (60.0) 104 (33.0) 21 (7.0)
Abbreviations: n, total number; NA, not applied; NK, not known. *Sporadic cases from south and central regions of Brazil. **In Real monthly ($R, the Brazilian currency).
predominant number of subjects with lower socio economic status, according to monthly family income (P ¼ 0.001) amongst cases and controls. There were no statistical differences among the distribution of AML subtypes. The frequency of AML-M3 observed in this series of cases was 16.4%, which is similar to the ratio reported elsewhere for acute promyelocytic leukemia in children [17]. The mean age was (7.1 + 5.8) for cases and (5.4 + 5.2) for controls [range 0.1 – 18 years]. The MTHFR C677T and A1298C genotypic distribution are in Hardy-Weinberg equilibrium in both case and control groups as shown in Table II. The frequencies obtained for the mutant alleles in the AML and control groups were 0.31 and 0.30 for MTHFR 677T allele and 0.26 and 0.23 for the MTHFR 1298C allele, respectively. The analysis of heterogeneity between cases and controls stratified by AML subtypes showed no significant differences, however statistic significance was found for the MTHFR A1298C genotypes distribution among non-white individuals (P ¼ 0.003).
*AML M0, M1, M6, M7 and AML not classified. **Borderline.
The distribution of MTHFR A1298C and C677T genotypes in AML cases and controls according to race is shown in Table III. No associations were found in the white children group in all analysis performed. In non-white children a protective effect for 677T allele was found for both heterozygotes and overall (CT þ TT) [0.37 OR, 95% CI, 0.14 – 0.92 and 0.45 OR, 95% CI, 0.20 – 1.02], respectively. In contrast, an increased risk of AML was observed for 1298C allele for heterozygotes [2.90 OR, 95% CI, 1.26 – 6.71] and overall [2.32 OR, 95% CI, 1.06 – 5.11]. The presence of the heterozygous combination [677 CT þ 1298 AC] was found in 9.3% and 11.4% of cases and controls, respectively. The combination of 677 TT homozygous þ 1298 AC heterozygous were observed only in 2.2% of cases (data not shown). When we applied the same logistic regression model to case-controls adjusted by race and geographical region, significant differences were maintained in the non-white children from the northeast [677 C 4 T ¼ 0.18 OR, 95% CI, 0.05 – 0.60] P ¼ 0.006, and [1298 A 4 C ¼ 7.61 OR, 95% CI, 2.69 – 21.49] P ¼ 0.0001; whereas in the southeast group there was no association risk for either white or non-white children. Considering speculations that the higher incidence of AML-M3 among Latinos in comparison to Europeans may be due to a different genetic background [17], we tested whether the MTHFR locus could be part of this genetic background. The subtypes M2 and M3 were selected, mainly because these groups are associated with the more common acquired chromosomal abnormalities in AML. The last is important in view of the interconnection of deficiencies in folic acid metabolism and the occurrence of chromosomal abnormalities. No significant
MTHFR polymorphisms in childhood AML associations were observed in this series of AML children as shown in Table IV.
2073
that the vast majority of AML subtypes are associated with a consistent variety of chromosomal translocations. These result in gene products that act as key elements in the disruption and block of normal cell differentiation [1]. Although it is assumed that genetic and environmental interactions described in ALL hold true for AML as well, very few studies were performed
Discussion AML, the most common acute leukemia in adults, is a relatively uncommon disease in childhood. In the few past years, several studies have demonstrated
Table III. Risk of AML associated with MTHFR C677T and A1298C genotypes according to race.
Whites
Polymorphism
MTHFR
C677T
CC CT TT CT þ TT AA AC CC AC þ CC
63 58 18 76 83 42 14 56
(46.3) (41.7) (12.9) (54.6) (59.7) (30.2) (10.0) (40.2)
102 84 23 107 117 80 12 92
(48.8) (40.1) (11.0) (51.1) (55.9) (38.2) (5.7) (44.0)
1.0* 1.12 (0.69 – 1.82) 1.27 (0.60 – 2.67) 1.15 (0.73 – 1.81) 1.0* 0.74 (0.45 – 1.21) 1.64 (0.67 – 4.03) 0.86 (0.54 – 1.36)
CC CT TT CT þ TT AA AC CC AC þ CC
30 9 4 13 21 20 2 22
(69.7) (20.9) (9.3) (30.2) (48.8) (46.5) (4.6) (51.1)
54 44 8 52 73 24 9 33
(50.9) (41.5) (7.5) (49.0) (68.8) (22.6) (8.4) (31.1)
1.0* 0.37 (0.14 – 0.92) 0.90 (0.21 – 3.69) 0.45 (0.20 – 1.02) 1.0* 2.90 (1.26 – 6.71) 0.77 (0.11 – 4.32) 2.32 (1.06 – 5.11)
CC CT TT CT þ TT AA AC CC AC þ CC
93 67 22 89 104 62 16 78
(51.1) (36.8) (12.0) (48.9) (57.1) (34.0) (8.7) (42.8)
156 128 31 159 190 104 21 125
(49.5) (40.6) (9.8) (50.4) (60.3) (33.0) (6.7) (39.7)
1.0* (0.58 – 1.32) (0.62 – 2.27) (0.64 – 1.38) 1.0* (0.72 – 1.65) (0.66 – 2.93) (0.77 – 1.68)
A1298C
Non-whites
C677T
A1298C
Total
C677T
A1298C
AML n (%)
Controls n (%)
OR (95% CI)
0.88 1.19 0.94 1.09 1.39 1.14
P-Yates
0.030 0.055 0.010 0.035
*Reference group (odds ratio ¼ 1.0). Table IV. Distribution of MTHFR genotypes in children and controls according to AML subtypes.
AML-M2
Polymorphism
MTHFR
C677T
CC CT TT CT þ TT AA AC CC AC þ CC
21 18 3 21 23 15 4 19
(50.0) (42.9) (7.1) (50.0) (54.8) (35.7) (9.5) (45.2)
156 128 31 159 190 104 21 125
(49.5) (40.6) (9.8) (50.4) (60.3) (30.0) (6.6) (39.6)
1.0* 1.04 (051 – 2.15) 0.72 (0.16 – 2.77**) 0.98 (0.49 – 1.96) 1.0* 1.19 (0.56 – 2.50) 1.57 (0.42 – 5.45**) 1.26 (0.63 – 2.51)
CC CT TT CT þ TT AA AC CC AC þ CC
13 12 5 17 18 10 2 12
(43.3) (40.0) (16.7) (56.7) (60.0) (33.3) (6.6) (39.9)
156 128 31 159 190 104 21 125
(49.5) (40.6) (9.8) (50.4) (60.3) (30.0) (6.6) (39.6)
1.0* 1.13 (0.46 – 2.73) 1.94 (0.56 – 6.42**) 1.28 (0.57 – 2.91) 1.0* 1.01 (0.42 – 2.42) 1.01 (0.0 – 5.00) 1.01 (0.44 – 2.31)
A1298C
AML-M3
C677T
A1298C
Abbreviations: OR, odds ratio; CI, confidence interval. *Reference group (OR ¼ 1.0). **Cornfield not accurate.
AML n (%)
Controls n (%)
OR (95% CI)
2074
F. J. Da Costa Ramos et al.
regarding the role of MTHFR polymorphisms in the AML pathogenesis [16]. This investigation demonstrated that the MTHFR 677 CT genotype was linked to a 2.7-fold decreased risk of developing AML amongst non-white children and young adults. In contrast, a significant 2.9-fold increased risk was observed for 1298 AC genotype. This finding corroborates recent data found by Zanrosso et al. in ALL children with the same demographic characteristics of this series [13]. On the other hand, these findings contrast with the data on the role of MTHFR A1298C in adult American patients, for whom a protective effect for this polymorphism was described [16]. The environmental-genetic linkage with respect to ALL emphasizing on the protective role of MTHFR C677T polymorphism has a plausible explanation. MTHFR-reduced activity in the C677T polymorphism increase the availability of 5,10methylene-tetrahydrofolate and intracellular folate metabolites required for DNA synthesis in normal cell divisions [18]. However, the MTHFR A1298C polymorphism is either associated with the higher homocysteine concentration or with the plasma folate interaction as observed in C677T polymorphism [11]. The MTHFR A1298C polymorphism lies in the S-adenosyl-methionine (SAM) regulatory domain of the enzyme resulting in conformational changes within the MTHFR enzyme that inhibit the enzyme’s activity. MTHFR C677T and MTHFR A1298C sites are 2.1 kb apart and have been found in strong linkage disequilibrium [19]. The combined MTHFR 677TT and 1298CC genotypes are extremely uncommon in the general population [11]. These findings suggest a founder effect in which the alteration evolved on a separate wild-type allele. In this study the combination of double heterozygous (677CT/1298AC) was found in 9.3% of cases and 11.43% of controls, the combination 677TT/ 1298AC was observed only in cases (2.2%) and both mutant homozygous 677TT/1298CC was not found. The founder effect in which altered combination evolved in these cases should be further evaluated in a large study. The heterogeneity of Brazilian population regarding racial definition mixed with social economic factors may represent a confounding factor herein. However, another study conducted in Bahia, a State in Northeast of Brazil, by Couto et al., that screened 843 neonates for MTHFR 677 demonstrated that the T677 allele frequency and TT677 genotype was higher than those observed in other studies of African-descent populations. Interesting was that MTHFR TT677 genotype was significantly more prevalent in children born in private maternity hospital (P ¼ 0.004) where the mothers who gave birth to these children are primarily representative of white-Brazilian population [20]. Our data related to
the allele frequencies were similar to the results of the northeast studies for the MTHFR CC677 and MTHFR AA1298 genotypes from previous reports [13,20]. Additional investigations combining demographic factors, environmental exposure data and AML molecular markers are necessary to provide more specific information regarding the chance of gene alterations and the interactions with cytogenetic abnormalities observed in AML cells. Acknowledgements This work was partially supported by grant-in-aid form INCA-FAF/Swiss Bridge Fund.[no. 230150-4]; MSPO has a CNPq-scholar grant no. 308532/2003-1; A.H. has a scholarship from CNPq. The authors are grateful to Programa Crianc¸a e Vida for the helpful support in the childhood cancer network care. We are also grateful to Ms Clarissa P. Oliveira for reviewing the manuscript. M.S.P.O and M.T.C.M contributed equally to this study, providing conception, design of the study, supervision and analysis of all data collected and, they also reviewed the manuscript; F.J.C.R., C.W.Z., V.C.S., M.A., M.P.M., performed techniques regarding molecular data and writing the manuscript; A.H., and E.M.F performed the statistical analysis and reviewed the manuscript; T.J.M., N.S., M.P.M., J.A.Y., S.R.B., helped in the assembly of cases and control samples and contributed with reviewing the manuscript. References 1. Tenen DG. Disruption of differentiation in human cancer: AML shows the way. Nature Reviews 2003;3(2):89 – 101. 2. Bhatia S, Neglia JP. Epidemiology of childhood acute myelogenous leukemia. J Pediatr Hematol Oncol 1995;17(2): 94 – 100. 3. Duthie SJ, Narayanan S, Brand GM, Pirie L, Grant G. Impact of folate deficiency on DNA stability. J Nutr 2002; 132(8 Suppl):2444S – 2449S. 4. Yamada K, Chen Z, Rozen K, Matthews RG. Effects of common polymorphisms on the properties of recombinant human methylenetetrahydrofolate reductase. PNAS 2001; 98(26):14853 – 14858. 5. Blount HC, Mack MM, Wehr CM, Macgregor JT, Hiatt RAW, Wang G, et al. Folate deficiency causes uracil misincorporation into human DNA and chromosome breakage: implications for cancer and neuronal damage. PNAS 1997;94:(7):3290 – 3295. 6. Wiemels JL, Smith RN, Taylor GM, Eden OB, Alexander FE, Greaves MF. Methylenetetrahydrofolate reductase (MTHFR) polymorphisms and risk of molecularly defined subtypes of childhood acute leukemia. PNAS 2001;98(7):4004 – 4009. 7. Franco RF, Simo˜es BP, Tone LG, Gabellini SM, Zago MA, Falca˜o RP. The methylenetetrahydrofolate reductase C677T gene polymorphism decreases the risk of childhood acute lymphocytic leukaemia. Br J Hematol 2001;115(3):616 – 618.
MTHFR polymorphisms in childhood AML 8. Oliveira E, Alves S, Quental S, Ferreira F, Norton L, Costa V, et al. The MTHFR C677T and A1298C polymorphisms and susceptibility to childhood acute lymphoblastic leukemia in Portugal. J Pediatr Hematol Oncol 2005;27(8):425 – 429. 9. Schnakenberg E, Mehles A, Cario G, Rehe K, Seidemann K, Schlegelberger B, et al. Polymorphisms of methylenetetrahydrofolate reductase (MTHFR) and susceptibility to pediatric acute lymphoblastic leukemia in a German study population. BMC Medical Genetics 2005;May 27:6(23). Available from the Alliance for Computers and Writing via the INTERNET 5http://www.biomedcentral.com/1471-2350/6/234 (Accessed 2005 May 14). 10. Chiusolo P, Reddiconto G, Cimino G, Sica S, Fiorini A, Farina G, et al. Methylenetetrahydrofolate reductase genotypes do not play a role in acute lymphoblastic leukemia pathogenesis in the Italian population. Haematologica 2004; 89(2):139 – 144. 11. Robien K, Ulrich CM. 5,10-methylenetetrahydrofolate reductase polymorphisms and leukemia risk: a HuGE Minireview. Am J Epidemiol 2003;157(7):571 – 582. 12. Bennett JM, Catovsky D, Daniel MT, Flandrin G, Gralcnik, Sultan C. Proposed revised criteria for the classification of the acute myeloid leukemia. A report of the French-AmericanBritish cooperative group. Ann Intern Med 1985;103(4): 620 – 625. 13. Zanrosso CW, Hatagima A, Emerenciano M, Ramos F, Figueiredo A, Felix TM, et al. The role of methylenetetrahydrofolate reductase in acute lymphoblastic leukemia in a Brazilian mixed population. Leuk Res 2006;30(4):477 – 481.
2075
14. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 1988;16(3):1215. 15. Frosst P, Blom HJ, Milos R, Goyette P, Sheppard CA, Matthews RG, et al. A candidate genetic risk factor for vascular disease: A common mutation at the methylenetetrahydrofolate reductase locus. Nature Gen 1995;10(1): 111 – 113. 16. Skibola CF, Smith MT, Kane E, Roman E, Rollinson S, Cartwright RA, et al. Polymorphisms in the methylenetetrahydrofolate reductase gene is associated with susceptibility to acute leukemia in adults. PNAS 1999;96(22):12216 – 12218. 17. Biondi A, Rovelli A, Cantu-Rajnoldi A, Fenu S, Basso G, Luciano A, et al. Acute promyelocytic leukemia in children: experience of the Italian Pediatric Hematology and Oncology Group (AIEOP). Leukemia 1994;8:1264 – 1268. 18. Bailey LB, Gregory JF 3rd. Polymorphisms of methylenetetrahydrofolate reductase and other enzyme: metabolic significance, risks and impact on folate requirement. J Nutr 1999;129(5):919 – 922. 19. Chen J, Ma J, Stampfer MJ, Palomeque C, Selhub J, Hunter DJ. Linkage disequilibrium between the C677T and 1298 A 4 C polymorphisms in human methylenetetrahydrofolate reductase gene and their contributions to risk of colorectal cancer. Pharmacogenetics 2002;12(4):339 – 342. 20. Couto FD, Adorno EV, Menezes JF, Moura Neto JP, Rego MA, dos Reis MG. C677T polymorphism of the MTHFR gene and variant haemoglobin: a study in newborns from Salvador, Bahia, Brazil. Cad Saude Publica 2004;20(2):529 – 533.
