Zurich Open Repository and Archive. Genetic variants of folate and methionine metabolism and PCNSL incidence in a German patient population

Transcription

1 University of Zurich Zurich Open Repository and Archive Winterthurerstr. 190 CH-8057 Zurich Year: 2010 Genetic variants of folate and methionine metabolism and PCNSL incidence in a German patient population Kurzwelly, D; Knop, S; Guenther, M; Loeffler, J; Korfel, A; Thiel, E; Hebart, H; Simon, M; Weller, M; Linnebank, M; Herrlinger, U Kurzwelly, D; Knop, S; Guenther, M; Loeffler, J; Korfel, A; Thiel, E; Hebart, H; Simon, M; Weller, M; Linnebank, M; Herrlinger, U (2010). Genetic variants of folate and methionine metabolism and PCNSL incidence in a German patient population. Journal of Neuro-Oncology, 100(2): Postprint available at: Posted at the Zurich Open Repository and Archive, University of Zurich. Originally published at: Kurzwelly, D; Knop, S; Guenther, M; Loeffler, J; Korfel, A; Thiel, E; Hebart, H; Simon, M; Weller, M; Linnebank, M; Herrlinger, U (2010). Genetic variants of folate and methionine metabolism and PCNSL incidence in a German patient population. Journal of Neuro-Oncology, 100(2):

2 Genetic variants of folate and methionine metabolism and PCNSL incidence in a German patient population Abstract Functional genetic polymorphisms involved in folate and methionine metabolism play an important role in both DNA synthesis and methylation, and affect the risk of various malignancies including lymphoproliferative disorders such as systemic non-hodgkin's lymphoma. In a retrospective analysis of 185 immunocompetent patients with primary central nervous system lymphoma (PCNSL) and 212 population controls we therefore investigated eight genetic polymorphisms affecting methionine metabolism for potential association with the development of PCNSL. We observed underrepresentation of the G-allele of the methyltetrahydrofolate homocysteine S-methyltransferase (MTR) c.2756a > G (D919G) missense polymorphism among PCNSL patients (P = 0.045; odds ratio (OR) = 0.65; ). Furthermore, for the methylenetetrahydrofolate reductase (MTHFR) c.1298a > C (E429A) polymorphism the mutated C-allele was found more frequently among PCNSL patients than among population controls (P = 0.026; OR = 1.57; ). There were no associations of the other polymorphisms investigated (MTHFR c.677c > T, transcobalamin 2 (Tc2) c.776c > G, cystathionin beta-synthase (CBS) c.844_855ins68, reduced folate carrier-1 (RFC-1) c.80g > A, thymidylate synthase (TYMS) 28-bp repeat, and dihydrofolate reductase (DHFR) c del19 bp) and the presence of PCNSL. This analysis is the largest to date to evaluate associations between genetic variants of folate and methionine metabolism and PCNSL. Our results suggest the hypothesis that folate and methionine metabolism is relevant to susceptibility to PCNSL.

3 Running head: Folate metabolism and PCNSL Genetic variants of folate and methionine metabolism and PCNSL incidence in a German patient population Delia Kurzwelly 1, Stefan Knop 2, Markus Guenther 3, Juergen Loeffler 2, Agnieszka Korfel 4, Eckhard Thiel 4, Holger Hebart 5, Matthias Simon 6, Michael Weller 7, Michael Linnebank 7, Ulrich Herrlinger 1 1 Division of Clinical Neurooncology, Department of Neurology, University of Bonn, Sigmund-Freud-Str. 25, D Bonn, Germany 2 Department of Hematology and Oncology, Wuerzburg University Hospital, Josef-Schneider- Straße 2, D Wuerzburg, Germany 3 Department of Internal Medicine, Klinikum Stuttgart, Prießnitzweg 24, Stuttgart, Germany 4 Department of Hematology and Oncology, Charité Campus Benjamin Franklin, Hindenburgdamm 30, D Berlin, Germany 5 Department of Internal Medicine, Klinikum Schwaebisch Gmuend Stauferklinik, Wetzgauer Straße 85, D Mutlangen, Germany 6 Department of Neurosurgery, University of Bonn, Sigmund-Freud-Str. 25, D Bonn, Germany 7 Department of Neurology, University Hospital Zurich, Frauenklinikstraße 26, CH-8091 Zurich, Switzerland

4 Correspondence to: Ulrich Herrlinger, MD, Division of Clinical Neurooncology, Department of Neurology, University of Bonn, Sigmund-Freud-Str. 25, D Bonn, Germany. Tel.: ; Fax: ; Abstract Functional genetic polymorphisms involved in folate and methionine metabolism play an important role in both DNA synthesis and methylation, and affect the risk of various malignancies including lymphoproliferative disorders such as systemic non-hodgkin s lymphoma. In a retrospective analysis of 185 immunocompetent patients with primary central nervous system lymphoma (PCNSL) and 212 population controls we therefore investigated eight genetic polymorphisms influencing methionine metabolism for a potential association with the development of PCNSL. We observed an underrepresentation of the G-allele of the methyltetrahydrofolate homocysteine S-methyltransferase (MTR) c.2756a>g (D919G) missense polymorphism among PCNSL patients (p=0.045; odds ratio [OR]=0.65; ). Furthermore, for the methylenetetrahydrofolate reductase (MTHFR) c.1298a>c (E429A) polymorphism the mutated C-allele was found more frequently among PCNSL patients in comparison to population controls (p=0.026; OR=1.57; ). There were no associations of the other polymorphisms investigated [MTHFR c.677c>t, transcobalamin 2 (Tc2) c.776c>g, cystathionin beta-synthase (CBS) c.844_855ins68, reduced folate carrier-1 (RFC-1) c.80g>a, thymidylate synthase (TYMS) 28-bp repeat, and dihydrofolate reductase (DHFR) c del19bp] and the presence of PCNSL. 2

5 The present analysis is the largest to date to evaluate associations between genetic variants of folate and methionine metabolism and PCNSL. Our results suggest the hypothesis that folate and methionine metabolism is relevant for the susceptibility to PCNSL. Keywords: Genetic polymorphism - Folate - Methionine - DNA methylation - PCNSL 3

6 Introduction Primary central nervous system lymphoma (PCNSL) is an aggressive subtype of non- Hodgkin s lymphoma (NHL) arising within the central nervous system (CNS). The vast majority are diffuse large B-cell lymphomas (DLBCL) derived from germinal center B cells [1], and pathogenic factors are largely unclear, so far. In a recent retrospective analysis no association between PCNSL and human leukocyte antigens (HLA) was found [2]. PCNSL characteristically reveals genetic instability with chromosomal imbalances [3]. Thus, metabolic factors and conditions contributing to the maintenance of DNA integrity may be influential for the development and growth of PCNSL cells via altered DNA synthesis and methylation of oncogenes and tumor suppressor genes. Aberrant folate and methionine metabolism could interfere with both processes. Genetic variants that functionally influence enzymes, transporter proteins or receptor proteins involved in folate and methionine metabolism are associated with different types of extracranial human cancer, such as systemic NHL, acute leukemia, and colorectal cancer [4-8]. In immunocompetent patients, an association of PCNSL with the methyltetrahydrofolate homocysteine S-methyltransferase (MTR) c.2756a>g variant has been reported [9], and this polymorphism has also been described to alter susceptibility to other intracranial tumors including glioblastoma and anaplastic meningioma [10, 11]. In the present explorative study we aimed to analyze the frequency of eight functional genetic variants of folate and methionine metabolism in a collective of 185 patients with PCNSL compared with 212 healthy controls, making this study the largest to date to evaluate this association. 4

7 Materials and methods Study population We investigated 185 consecutive immunocompetent PCNSL patients of Caucasian origin recruited for the German multicenter phase IV trial (G-PCNSL-SG-1) between 08/2000 and 12/2004 (43.8% female; median age at diagnosis 60.4 years, range years). The aim of the G-PCNSL-SG-1 study is to analyze the value of whole brain radiotherapy after 6 courses of high-dose methotrexate (MTX) in newly diagnosed PCNSL. Inclusion criteria for the G- PCNSL-SG-1 study were a newly diagnosed and histologically or cytologically (in the cerebrospinal fluid) confirmed PCNSL in an immunocompetent patient and adequate renal and bone marrow function. Exclusion criteria were systemic manifestation of NHL and additional malignancies as well as immunodeficiency or concomitant immunosuppressive therapy. 212 apparently healthy Caucasian Bonn area residents (45.8% female; median age 63.0 years, range years) without a history of cancer, recruited as control population for an ongoing study on atherosclerosis, served as controls [12, 13]. The same collective had already been used as control group for a small case-control study analyzing folatemetabolizing pathway polymorphisms in patients with PCNSL [9]. Characteristics of the study population have also been summarized in Table 1. The study was approved by the respective local ethics committees, and all participants gave written informed consent. DNA extraction and genotyping Blood samples were collected prospectively upon enrollment onto the G-PCNSL-SG-1 trial before treatment was started. Genomic DNA was extracted from mononuclear cells using the QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany). Genotyping of eight polymorphisms [methylenetetrahydrofolate reductase (MTHFR) c.677c>t (A222V), MTHFR c.1298a>c (E429A), MTR c.2756a>g (D919G), transcobalamin 2 (Tc2) c.776c>g (P259R), 5

8 cystathionin beta-synthase (CBS) c.844_855ins68 (change of transcript levels), reduced folate carrier-1 (RFC-1) c.g80a (R27H), thymidylate synthase (TYMS) 28bp rep (2R>3R), dihydrofolate reductase (DHFR) c del19bp (change of transcript levels)] was performed by amplification of genomic DNA applying polymerase chain reaction (PCR) and subsequent restriction enzyme digestion or by allele specific PCR, followed by agarose gel electrophoresis [14-20]. Statistical analysis The distribution of age and gender in the patient and the control collective was compared applying t-test for two independent samples and Pearson`s Chi 2 test, respectively. For the tested genotypes the Hardy-Weinberg equation was calculated for all PCNSL patients and controls together with a Chi 2 goodness-of-fit test (df=2). The two-sided Pearson`s Chi 2 test was used to analyze the distribution of the respective genotypes in the patient and the control group for statistical significance. Threshold was defined with alpha <0.05. Due to the low frequencies of the MTR c.2756gg and the MTHFR c.1298cc homozygous variants, we additionally evaluated the presence of at least one mutant allele (AG/GG or AC/CC) versus homozygosity for the wildtype allele (AA) in the patient and the control sample (Pearson`s Chi 2 test; df=1). All analyses were purely explorative, so that correction for multiple testing was not regarded necessary, and all findings require reevaluation in future studies. Odds ratios (OR) together with 95% confidence intervals (CI) were calculated for the combined MTR c.2756ag/gg and MTHFR c.1298ac/cc genotypes, respectively. Furthermore, multinominal regression analysis with alpha=0.05 was applied to test the independent association of the respective polymorphisms with PCNSL and to exclude confounding effects of age, gender, or multiple testing. All statistical calculations were performed using the Statistical Package for the Social Sciences ( SPSS ) software version 16.0 (SPSS, Chicago, IL, USA). 6

9 Results There were no significant differences concerning age and gender between the patient and the control cohort. Genotype distributions in all PCNSL patients and controls are shown in Table 2. The distribution of allelotypes of the polymorphisms did not significantly deviate from the Hardy-Weinberg equilibrium (data not shown). Allelic frequencies of the MTHFR c.677c>t, MTHFR c.1298a>c, MTR c.2756a>g, Tc2 c.776c>g, CBS c.844_855ins68, RFC-1 c.80g>a, TYMS 28-bp repeat, and the DHFR c del19bp polymorphism in the control group are similar to those found in another independent healthy German population [21] and those reported for other Caucasian populations [15, 16, 22]. The MTR c.2756a>g polymorphism (G-allele) was less frequent in PCNSL patients than in controls (patients AA/AG/GG: 71.3/24.9/3.8 and controls AA/AG/GG: 61.8/34.0/4.2; Chi 2 =4,17 (Pearson), p=0.125 for trend, df=2). In Pearson`s Chi 2 test of the pooled genotypes MTR c.2756aa versus AG/GG without correction for multiple testing, the MTR c.2756g allele was significantly underrepresented in the patient sample; at least one mutant allele (the MTR c.2756ag and GG genotypes) was detected in 28.7% of PCNSL patients compared to 38.2% of controls (Chi 2 =4.04, p=0.045, df=1; Table 3). Overall, multinominal logistic regression analysis with simultaneous analysis of all eight polymorphisms together with age and gender as co-variables reproduced the above findings (Chi 2 =4.449, p=0.035 for pooled genotypes, df=1). Furthermore, for the MTHFR c.1298a>c polymorphism, the mutated C-allele was observed more frequently among PCNSL patients than in controls (patients AA/AC/CC: 38.8/52.1/9.1 and controls AA/AC/CC: 50.5/42.0/7.5; Chi 2 =5.38 (Pearson), p=0.068 for trend, df=2). This trend was significant in the uncorrected explorative analysis (MTHFR c.1298aa versus AC 7

10 and CC genotypes: Chi 2 =5.00, p=0.026, df=1; Table 3) as well as in multinomial logistic regression analysis of the pooled genotypes (Chi 2 =5.179, p=0.023). A reduced risk estimate was found for the MTR c.2756ag/gg genotype when MTR c.2756aa was the referent group (OR=0.65; ; Table 3), whereas a small increased risk estimate for PCNSL was associated with the MTHFR c.1298ac/cc genotype versus homozygosity for the wildtype allele AA (OR=1.57; ; Table 3). There were no significant associations of the other polymorphisms investigated (MTHFR c.677c>t, Tc2 c.776c>g, CBS c.844_855ins68, RFC-1 c.80g>a, TYMS 28-bp repeat, and DHFR c del19bp; Table 2) and the development of PCNSL. Discussion The findings presented here suggest an association between the MTR c.2756a>g (D919G) as well as the MTHFR c.1298a>c (E429A) polymorphisms and the presence of PCNSL, with the MTR c.2756g allele being less frequent, and the mutated MTHFR c.1298c allele more common in patients than in population controls. As for the MTR c.2756a>g polymorphism these data confirm the results of a previous casecontrol study on 31 German PCNSL patients which found significantly fewer carriers of the MTR c.2756a>g missense polymorphism among the patients [9]. Of note, control samples of both studies overlap, so the observed effect could have been due to a coincidental overrepresentation of the G-allele in the control group. However, this seems unlikely, as allelic frequencies of the MTR c.2756a>g polymorphism in this control group are similar to those of another independent healthy German population [21]. Interestingly, in this population 8

11 sample the presence of the MTR c.2756g allele has been correlated with disease-free longevity. MTR catalyzes the remethylation of homocysteine to methionine. The G-allele of MTR c.2756a>g has been suggested to lead to a higher MTR activity, supposedly changing the kinetics of nucleic acid and S-adenosylmethionine (SAM) synthesis. SAM is necessary for DNA methylation. Thus, the MTR variant may influence both DNA synthesis and methylation, whereas any concrete influence of the variant on tumorigenesis remains speculative. Our results showing a reduced frequency of the MTR c.2756g allele in patients with PCNSL are in line with previous studies on the MTR missense dimorphism c.2756a>g which found a lower frequency in patients with colorectal cancer [7], cervical intraepithelial neoplasia [23] and follicular NHL [8], and (among primary brain tumors) with GBM and anaplastic meningioma [10, 11]. The second genetic variant in this study, MTHFR c.1298a>c, with a significant positive association with PCNSL in explorative data analysis has previously been reported to be associated with some types of human cancer [4-6, 24]. MTHFR catalyzes the conversion of 5,10-methylenetetrahydrofolate to 5-methyltetrahydrofolate, which acts as a methyl donor for the transformation of homocysteine to methionine. Both polymorphisms in the MTHFR gene, c.677c>t and c.1298a>c, lead to reduced enzyme activity. While these variants were both observed more frequently in our patient group, this effect did not reach statistical significance for the MTHFR c.677c>t polymorphism. That could be due to a too small sample size and should be reevaluated in further patient collectives, although PCNSL is a rare disease, and large patient cohorts are difficult to obtain. Possible mechanisms underlying an association of the MTHFR variants include effects on folate derivatization, on homocysteine remethylation to methionine and on SAM availability for DNA synthesis. More basic research is necessary to elucidate such putative interactions. 9

12 However, as pointed out, the reported variants do not exclusively affect pathways in PCNSL development but are apparently relevant for tumorigenesis in general. Moreover, due to the case-control design of our study, and due to the "matching procedure" we employed (which was not done between individual cases and controls, but between the entire group of cases and the entire group of controls) certain caveats are necessary, including the possibility of unrecognized selection bias and unknown confounders which cannot be fully controlled for. As a result, a definitive relationship between "exposure" (i.e. MTR and MTHFR polymorphisms) and PCNSL has not been established. This study is an explorative, hypothesis-generating investigation. Our results require replication in additional large trials and in pooled analyses. In summary, the present analysis is the largest to date to evaluate associations between polymorphisms in folate-metabolizing genes and PCNSL, and suggests that certain functional genetic variants of folate and methionine metabolism alter susceptibility to PCNSL. Specifically, the MTR c.2756g allele may reduce susceptibility (a protective effect) while the MTHFR c.1298c allele may increase susceptibility. Since folate and methionine metabolism is closely associated with the availability of methionine, folate and vitamin B12, it might even be possible that nutritional factors modify the incidence of PCNSL, as has been described for other tumors. Folate and methionine metabolism also seems to affect anticancer therapy with MTX which is still the most efficient chemotherapy component in the treatment of PCNSL and directly interferes with methionine metabolism. Therefore, further studies on the impact of folate and methionine metabolism on incidence and treatment outcome of PCNSL are warranted. 10

13 Acknowledgment: This study was generously supported by a grant from the Fortune Program of the University of Tuebingen Medical School to S. Knop, H. Hebart, and U. Herrlinger and a Deutsche Krebshilfe grant ( Th2) to U. Herrlinger, A. Korfel, E. Thiel, and M. Weller. References 1. Montesinos-Rongen M, Küppers R, Schlüter D, Spieker T, Van Roost D, Schaller C, Reifenberger G, Wiestler OD, Deckert-Schlüter M (1999) Primary central nervous system lymphomas are derived from germinal-center B cells and show a preferential usage of the V4-34 gene segment. Am J Pathol 155: Kurzwelly D, Müller CA, Korfel A, Thiel E, Linnebank M, Weller M, Herrlinger U (2008) Primary CNS lymphoma and HLA class I and II alleles in a German cohort of immunocompetent patients. J Neurooncol 90: Weber T, Weber RG, Kaulich K, Actor B, Meyer-Puttlitz B, Lampel S, Büschges R, Weigel R, Deckert-Schlüter M, Schmiedek P, Reifenberger G, Lichter P (2000) Characteristic chromosomal imbalances in primary central nervous system lymphomas of the diffuse large B-cell type. Brain Pathol 10: Sharp L, Little J (2004) Polymorphisms in genes involved in folate metabolism and colorectal neoplasia: a HuGE review. Am J Epidemiol 159: Skibola CF, Smith MT, Kane E, Roman E, Rollinson S, Cartwright RA, Morgan G (1999) Polymorphisms in the methylenetetrahydrofolate reductase gene are associated with susceptibility to acute leukemia in adults. Proc Natl Acad Sci USA 96: Matsuo K, Suzuki R, Hamajima N, Ogura M, Kagami Y, Taji H, Kondoh E, Maeda S, Asakura S, Kaba S, Nakamura S, Seto M, Morishima Y, Tajima K (2001) Association 11

14 between polymorphisms of folate- and methionine-metabolizing enzymes and susceptibility to malignant lymphoma. Blood 97: Ma J, Stampfer MJ, Christensen B, Giovannucci E, Hunter DJ, Chen J, Willett WC, Selhub J, Hennekens CH, Gravel R, Rozen R (1999) A polymorphism of the methionine synthase gene: association with plasma folate, vitamin B12, homocysteine, and colorectal cancer risk. Cancer Epidemiol Biomarkers Prev 8: Lincz LF, Scorgie FE, Kerridge I, Potts R, Spencer A, Enno A (2003) Methionine synthase genetic polymorphism MS A2756G alters susceptibility to follicular but not diffuse large B-cell non-hodgkin`s lymphoma or multiple myeloma. Br J Haematol 120: Linnebank M, Schmidt S, Kölsch H, Linnebank A, Heun R, Schmidt-Wolf IG, Glasmacher A, Fliessbach K, Klockgether T, Schlegel U, Pels H (2004) The methionine synthase polymorphism D919G alters susceptibility to primary central nervous system lymphoma. Br J Cancer 90: Semmler A, Simon M, Moskau S, Linnebank M (2006) The methionine synthase polymorphism c.2756a>g alters susceptibility to glioblastoma multiforme. Cancer Epidemiol Biomarkers Prev 15: Semmler A, Simon M, Moskau S, Linnebank M (2008) Polymorphisms of methionine metabolism and susceptibility to meningioma formation: laboratory investigation. J Neurosurg 108: Moskau S, Golla A, Grothe C, Boes M, Pohl C, Klockgether T (2005) Heritability of carotid artery atherosclerotic lesions: an ultrasound study in 154 families. Stroke 36: Linnebank M, Moskau S, Farmand S, Fliessbach K, Kölsch H, Bös M, Grothe C, Becker D, Harbrecht U, Pohl C, Wüllner U, Klockgether T (2006) Homocysteine and carotid 12

15 intima-media thickness in a German population: Lack of clinical relevance. Stroke 37: Afman LA, Lievers KJ, van der Put NM, Trijbels FJ, Blom HJ (2002) Single nucleotide polymorphisms in the transcobalamin gene: relationship with transcobalamin concentrations and risk for neural tube defects. Eur J Hum Genet 10: Harmon DL, Shields DC, Woodside JV, McMaster D, Yarnell JW, Young IS, Peng K, Shane B, Evans AE, Whitehead AS (1999) Methionine synthase D919G polymorphism is a significant but modest determinant of circulating homocysteine concentrations. Genet Epidemiol 17: Linnebank M, Homberger A, Junker R, Nowak-Goettl U, Harms E, Koch HG (2001) High prevalence of the I278T mutation of the human cystathionine ß-synthase detected by a novel screening application. Thromb Haemost 85: Johnson WG, Stenroos ES, Spychala JR, Chatkupt S, Ming SX, Buyske S (2004) New 19 bp deletion polymorphism in intron-1 of dihydrofolate reductase (DHFR): a risk factor for spina bifida acting in mothers during pregnancy? Am J Med Genet A 124: Winkelmayer WC, Eberle C, Sunder-Plassmann G, Fodinger M (2003) Effects of the glutamate carboxypeptidase II (GCP2 1561C>T) and the reduced folate carrier (RFC1 80G>A) allelic variants on folate and total homocysteine levels in kidney transplant patients. Kidney Int 63: Frosst P, Blom HJ, Milos R, Goyette P, Sheppard CA, Matthews RG, Boers GJH, den Heijer M, Kluijtmans LAJ, van den Heuve LP, Rozen R (1995) A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase. Nat Genet 10: van der Put NM, Gabreëls F, Stevens EM, Smeitink JA, Trijbels FJ, Eskes TK, van den Heuvel LP, Blom HJ (1998) A second common mutation in the 13

16 methylenetetrahydrofolate reductase gene: an additional risk factor for neural-tube defects? Am J Hum Genet 62: Linnebank M, Fliessbach K, Kolsch H, Rietschel M, Wullner U (2005) The methionine synthase polymorphism c.2756a>g (D919G) is relevant for disease-free longevity. Int J Mol Med 16: Linnebank M, Homberger A, Nowak-Gottl U, Marquardt T, Harms E, Koch HG (2000) Linkage disequilibrium of the common mutations 677C>T and 1298A>C of the human methylenetetrahydrofolate reductase gene as proven by the novel polymorphisms 129C>T, 1068C>T. Eur J Pediatr 159: Henao OL, Piyathilake CJ, Waterbor JW, Funkhouser E, Johanning GL, Heimburger DC, Partridge EE (2005) Women with polymorphisms of methylenetetrahydrofolate reductase (MTHFR) and methionine synthase (MS) are less likely to have cervical intraepithelial neoplasia (CIN) 2 or 3. Int J Cancer 113: Wiemels JL, Smith RN, Taylor GM, Eden OB, Alexander FE, Greaves MF; United Kingdom Childhood Cancer Study investigators (2001) Methylenetetrahydrofolate reductase (MTHFR) polymorphisms and risk of molecularly defined subtypes of childhood acute leukemia. Proc Natl Acad Sci USA 98:

17 Table 1 Characteristics of patients with primary central nervous system lymphoma (PCNSL) and control individuals Characteristics PCNSL Controls No. (%) Median (Range) No. (%) Median (Range) No. of individuals Demographic data Male 104 (56.2) (54.2) - Female 81 (43.8) - 97 (45.8) - Age at diagnosis [years] (18-81) (39-85) Descent Caucasian Caucasian Prior therapy None -

18 Table 3 Pearson`s Chi 2 analysis, odds ratios (OR), and 95% confidence intervals (CI) of the pooled genotypes MTR c.2756ag/gg vs. AA and MTHFR c.1298ac/cc vs. AA in PCNSL patients and controls Genotype PCNSL (n = 185) Controls (n = 212) MTR c.2756a>g χ 2 (df = 1) P (χ 2 ) OR (95% CI) AA AG or GG ( ) MTHFR c.1298a>c χ 2 P OR (95% CI) AA AC or CC ( ) The distribution of the different genotypes among PCNSL patients and controls is given as relative amount.

19 Table 2 Distribution of polymorphisms involved in folate-metabolizing pathway genes in patients with primary central nervous system lymphoma (PCNSL) and controls: Pearson`s Chi 2 analysis MTHFR c.677c>t CC CT TT χ 2 (df = 2) P (χ 2 ) PCNSL (n = 185) Controls (n = 212) MTHFR c.1298a>c AA AC CC χ 2 P PCNSL (n = 185) Controls (n = 212) RFC c.80g>a GG GA AA χ 2 P PCNSL (n = 185) Controls (n = 212) TYMS 28-bp repeat (2R>3R) 2R / 2R 2R / 3R 3R / 3R χ 2 P PCNSL (n = 185) Controls (n = 212) MTR c.2756a>g AA AG GG χ 2 P PCNSL (n = 185) Controls (n = 212) DHFR c del19-bp ins / ins ins / del del / del χ 2 P PCNSL (n = 185) Controls (n = 212) CBS c.844_855ins68 del /del del / ins ins / ins χ 2 P PCNSL (n = 185) Controls (n = 212) Tc2 c.776c>g CC CG GG χ 2 P PCNSL (n = 185) Controls (n = 212)

20 The rate of carriers of the different genotypes is given as relative amount. The Chi 2 test values for Pearson`s Chi 2 test for all three genotypes (= two degrees of freedom) for the differences between patients and controls are shown.