Genetic Susceptibility to Lymphoma

Abstract

BACKGROUND

Genetic susceptibility studies of lymphoma may serve to identify at risk populations and to elucidate important disease mechanisms.

METHODS

This review considered all studies published through October 2006 on the contribution of genetic polymorphisms in the risk of lymphoma.

RESULTS

Numerous studies implicate the role of genetic variants that promote B-cell survival and growth with increased risk of lymphoma. Several reports including a large pooled study by InterLymph, an international consortium of non-Hodgkin lymphoma (NHL) case-control studies, found positive associations between variant alleles in TNF -308G>A and IL10 -3575T>A genes and risk of diffuse large B-cell lymphoma. Four studies reported positive associations between a GSTT1 deletion and risk of Hodgkin and non-Hodgkin lymphoma. Genetic studies of folate-metabolizing genes implicate folate in NHL risk, but further studies that include folate and alcohol assessments are needed. Links between NHL and genes involved in energy regulation and hormone production and metabolism may provide insights into novel mechanisms implicating neuro- and endocrine-immune cross-talk with lymphomagenesis, but will need replication in larger populations.

CONCLUSIONS

Numerous studies suggest that common genetic variants with low penetrance influence lymphoma risk, though replication studies will be needed to eliminate false positive associations.

Introduction

Non-Hodgkin lymphoma (NHL) is a heterogeneous malignancy of B- and T-cells that involves their uncontrolled clonal expansion in the periphery. B-cell lymphomas comprise the majority of cases and, of these, diffuse large B-cell lymphoma (DLCL) and follicular lymphoma (FL) are the two major subtypes (1). In 2007, NHL will account for approximately 59,000 newly diagnosed cases and 19,000 deaths in the U.S. and over 300,000 cases and 172,000 deaths worldwide. New therapeutic regimens have begun to delay the number of deaths related to NHL, though causes for most cases remain undetermined. Familial aggregations for lymphoproliferative cancers have been documented for lymphomas and leukemias (2),(3),(4), and several large-scale studies have reported associations between family history of hematopoietic malignancy and lymphoma risk (5),(6),(7),(8),(9),(10). However, twin studies do not support the role of highly penetrant genes in NHL risk. It is likely that more common genetic variants, each with apparently minor effects on tumor phenotype, may influence disease susceptibility with a greater population attributable risk. Their additive or multiplicative effects and interactions with environmental and infectious agents likely play a significant role in disease risk, but have yet to be estimated due to the large populations required to measure these effects.

To date, a number of case-control association studies have examined the role of genetic polymorphisms in the risk of lymphoma and several genetic variants have been identified as potential susceptibility loci. These studies may serve to identify at risk populations and to further elucidate important disease mechanisms. In this review, we aim to summarize results of existing genetic association studies of lymphoma, discuss their potential biological relevance and propose possible areas for future studies. Given the importance of understanding the underlying mechanisms involved in the pathogenesis of lymphoma, first we will briefly describe characteristics of genetic instability that occurs as part of normal B-cell development that may lead to pre-neoplastic changes relevant to lymphoma.

Normal B-cell development and genetic instability: potential underlying mechanisms of lymphomagenesis

B-cells differentiate from hematopoietic stem cells within the bone marrow and their maturation occurs in several stages involving genetic recombination and mutation to generate high affinity antibodies. Recombinase activating gene proteins initiate DNA double-strand breaks allowing a fixed set of gene segments [variable, diversity, joining (VDJ)] to recombine to produce B-cell receptors specific for antigen. During this process, chromosomal translocations and mutations involving the immunoglobulin heavy (IgH) and light (IgL) chain genes and the T-cell receptor genes are common that can deregulate several oncogenes (11). Further translocations and mutations can occur following antigenic dependent clonal expansion in the lymph nodes involving somatic hypermutation and class switching. These translocations and somatic mutations can disrupt B-cell homeostasis and lead to proliferation, blocked differentiation and immortalization of B-cells. Genetic factors that impair DNA repair can increase the likelihood of these pre-neoplastic lesions. Genomic lesions in B-cells that are not initially lethal could later be modulated by environmental (i.e., infectious agents), epigenetic (i.e., chromosomal hypo-or hypermethylation), disease (i.e., autoimmune disease) and genetic factors (i.e., genetic polymorphisms) that promote B-cell survival and proliferation, and may lead to the development and progression of lymphoma (Figure 1). Below we discuss how genetic polymorphisms may play a contributory role in this process.

Figure 1. The normal life cycle of B-lymphocytes and the derivation of lymphoma subtypes.

During normal B-cell development, hematopoietic stem cells first colonize the bone marrow and give rise to common lymphoid progenitor cells, some of which will differentiate to B-cell lineage. While in the bone marrow, V(D)J recombination machinery rearranges germline immunoglobulin (Ig) gene loci that lead to formation of chromosomal translocations. Mature naïve B-cells that express the B-cell receptor exit the bone marrow to lymph nodes and extra-lymphatic follicles. There, upon antigenic stimulation, B-cells become activated undergoing a proliferative burst, generating formation of germinal centers (GCs). In GCs, proliferating B-cells are subjected to somatic hypermutation, directed at Ig genes. In GCs, the most common aggressive lymphoma, diffuse large B-cell lymphoma (DLCL), can originate from activated B-cells, otherwise known as centroblasts, and the most common indolent lymphoma, follicular lymphoma (FL), can derive from centrocytes. FL also can transform to DLCL. Burkitt's lymphoma can derive from IgM-positive blasts of the early GC reaction. Some B-cells in GCs will further differentiate into memory cells while others will become plasma cells, of which multiple myeloma can derive. Mantle zone lymphomas and some small lymphocytic lymphomas have unmutated V-region genes suggesting they originate from naïve peripheral B-cells. Class switch recombination permits B-cells to switch from membrane-bound to soluble B-cell receptors and also acts aberrantly to cause chromosomal translocations.

Candidate Gene Association Studies of Lymphoma

The majority of candidate genes that have been investigated to date can be categorized into “functional groups” based on their potential biological relevance. One group includes genes that influence DNA integrity and methylation. Polymorphisms in these genes could modulate the rate of chromosomal translocations, efficiency of DNA repair and DNA methylation status. Another major group involves genes that alter B-cell survival and growth including pro-inflammatory and regulatory cytokine genes, and genes involved in innate immunity, oxidative stress, energy regulation and hormone production. Another group are genes involved in xenobiotic metabolism. Risk alleles found in these pathways may provide clues to identify potential lymphomagens. On the basis of these studies, several gene variants have been associated with NHL; some have been replicated in more than one study, suggesting their significance in lymphomagenesis. A summary of results from these studies is listed in Table 1 with particular emphasis on those that reported positive findings.

Table 1. Summary of genetic polymorphisms investigated in lymphoma case-control and cohort studies.

If not otherwise indicated, odds ratios (OR) and 95% confidence intervals (in brackets) are given for the association of the rare allele carrier versus the more common allele homozygous genotype. OR with confidence intervals excluding the 1 and OR of borderline significance have been included. Subtype-specific results are indicated when available and noteworthy.

Pathway	Gene	Allele		Association with Lymphoma	Study population	Reference
			rs number		Cases/controls
DNA Repair
NHEJ	LIG4	Tyr9Ile	1805388	TT vs CC: OR=0.5 (0.3-0.9) for NHL, effect seen for FL, DLCL	1172 NHL/983	(13)
	RAG1	Lys820Arg	2227973	AG vs AA: OR=1.3 (1.0-1.6), GG vs AA: OR=2.7 (1.4-5.0) for NHL; effect seen for FL	1172 NHL/983	(13)
DSBR	WRN	Cys1367Arg	1346044	OR=0.71 (0.56-0.91) for NHL, effect seen for DLCL, FL, MZL	518 NHL/597	(12)
		Val114Ile	4987236	OR=0.7 (0.6-1.0) for NHL	1172 NHL/983	(13)
	BRCA1	Glu997Gly	16941	OR=0.79 (0.62-1.02) for NHL, effect seen for DLCL	518 NHL/597	(12)
	BRCA2	Asn289His	766173	OR=3.97 (1.60-9.90) for T-cell NHL	518 NHL/597	(12)
		Asn372His	144848	CC vs AA: OR=1.5 (1.0-2.1) for NHL, OR=2.0 (1.2-3.3) for T-cell NHL	1172 NHL/983	(13)
	XRCC3	Thr241Met	861539	OR=1.62 (1.03-2.56) for FL; OR=2.60 (1.04-6.51) for MZL	518 NHL/597	(12)
		UTR C>T	3212024	OR=1.8 (1.1-2.8) for FL	430 FL/605	(14)
		UTR T>C	3212038	OR=1.5 (1.0-2.4) for FL	430 FL/ 605	(14)
		Intron 4 G>A	3212090	OR=1.5 (1.0-2.5) for FL	430 FL/605	(14)
	NBS	657del		OR=8.05 (1.71-38.0)	42 NHL, 30 HL/1289	(113)
MR	TP53	Arg72Pro	1042522	OR=1.59 (0.99-2.57) for NHL	103 NHL/440	(114)
	MSH2	gIVS12 −6T->C	2303428	C: 0.11% in NHL, 0.05% in Co, p<0.01	22 NHL/50	(16)
				OR=1.44 (0.94-2.23) for NHL	103 NHL/487	(17)
	XRCC1	Arg194Trp	1799782	OR=0.45 (0.21-0.95) for FL	518 NHL/597	(12)
				OR=1.4 (1.1-1.8) for NHL, effect seen for B- and T-NHL	1172 NHL/983	(13)
		Arg280His	25489	OR=2.66 (1.35-5.26) for T-NHL, OR=2.01 (1.11-3.63) for SLL/CLL	518 NHL/597	(12)
NER	ERCC5	Asp1104His	17655	OR=1.46 (1.13-1.88) for NHL	518 NHL/597	(12)
	ERCC2	IVS5 −282G>A	1618536	GA vs GG: OR=1.5 (1.1-2.1) for FL, AA vs GG: NS	430 FL/605	(14)
		IVS16 137C>T	2070831	CT vs TT: OR=1.9 (1.1-3.2) for FL, TT vs CC: NS	430 FL/605	(14)
		Lys751Gln	13181	OR=0.64 (0.44-0.92) for DLCL, OR=0.56 (0.32-0.99) for SLL/CLL	518 NHL/597	(12)
One Carbon Metabolism	MTHFR	677C>T	1801133	OR=0.64 (0.39-1.05) for NHL	98 NHL/243	(34)
				OR=1.5 (0.97-2.2) for NHL; OR=1.8 (1.0-3.1) for FL, OR=1.8 (0.96-3.2) for DLCL	337 NHL/731	(22)
				CT vs TT: OR=0.59 (0.44-0.81) for NHL; CC vs TT: OR=0.81 (0.55-1.20) for NHL	372 NHL, HL/500	(24)
		1298A>C	1801131	AC vs AA: OR=1.26 (0.96-1.65) for NHL, CC vs AA: OR=1.27 (0.70-2.31) for NHL	372 NHL, HL/500	(24)
	MTR	Asp919Gly	1805087	GG: OR=3.59 (1.11-11.5) for NHL	98 NHL/243	(34)
				OR=0.41 (0.19-0.88) for FL, NS for DLCL	151 NHL, 90 MM/299	(28)
				OR=1.3 (0.99-1.7) for NHL, OR=1.3 (0.88-2.0) for DLCL, OR=1.2 (0.78-1.7) for FL	337 NHL/771	(22)
				OR=0.26 (0.09-0.74) for PCNSL	31 PCNSL/142	(29)
				GG: OR=2.00 (1.05-3.80) for NHL	372 NHL, HL/500	(24)
				OR=0.49 (0.29-0.82) for FL	172 FL/206	(30)
	MTRR	66A>G	1801394	A-allele: OR=0.69 (0.50-0.96) for NHL	120 ALL*, 200 NHL/257	(25)
	TYMS	TYMS 28-bp 2R> 3R		2R (+) vs 2R (−): OR=1.63 (1.05-2.53) for NHL	108 NHL/494	(33)
				2R3R: OR=1.48 (1.12-1.97) for NHL; OR=1.45 (0.96-2.2) for FL; OR=3.38 (1.30-8.82) for MZL, NS for DLCL	589 NHL/755	(23)
				2R (−) vs 2R (+): OR=2.12 (1.32-3.4) for FL	172 FL/206	(30)
		1494del6 TYMS IVS6 - 68C>T TYMS 1053C>T TYMS 1122A>G	16430, 1059394, 699517, 2790	6bp(−)(−), IVS6 −68TT, 1053TT genotypes (all in complete LD): OR=0.57 (0.34-0.94) for NHL, OR=0.29 (0.10-0.82) for DLCL, OR=0.67 (0.32-1.4) for FL	337 NHL/731	(22)
		1494del6	16430	Homozygous 6-bp deletion: OR=1.61 (0.99-2.60) for DLCL, NS for NHL and FL	589 NHL/755	(23)
	SHMT1	1420C>T	1979277	OR=0.46 (0.23-0.93) for NHL	108 NHL/494	(33)
				TT vs. CC: OR=0.75 (0.51-1.09) for NHL	589 NHL/755	(23)
	RFC	80G>A	1051266	NS for NHL and DLCL, 80AA: OR=1.44 (0.94-2.22) for FL	589 NHL/755	(23)
	CBS	844ins68 (splice) variant		OR=2.43 (0.90-6.61) for PCNSL	31 PCNSL/142	(29)
Immune regulation
Cytokines	TNF	−308G>A	1800629	A-allele in NHL Ca: OR=6.7, p<=0.0001	44 NHL/106	(115)
				OR=3.18 (1.57-8.3) for CLL	73 CLL	(116)
				OR=1.19 (1.05-1.33) for NHL, GA vs GG: OR=1.29 (1.10-1.51) for DLCL, AA vs GG: OR=1.65 (1.16-2.34)] for DLCL, NS for FL	3586 NHL/ 4018	(37)
				AA vs GG: OR=3.63 (p=0.028) for NHL	194 NHL/160	(117)
		−857C>T	2507961	OR=2.34 (1.2-4.7) for NHL	71 ATL, 80 healthy HTLVI	(53)
				OR=0.33 (0.15-0.75) for NHL	70 maltoma/ 210	(52)
				OR=1.8 (1.1-2.8) for MALT compared to H. pylori positive (HP +) controls	144 GMALT/594 HP+	(51)
				OR=1.9 (1.0-3.6) for MALT compared to blood donors
				OR=0.59 (0.42-0.84) for NHL, OR=0.40 (0.23-0.68) for FL	545 NHL/498	(50)
		−863C>A	1800630	OR=1.53 (1.03-2.27) for DLCL	545 NHL/498	(50)
	LTA	252 A>G	909253	OR=0.59 (0.37-1.04) for FL	121 FL/88	(118)
				252GG: OR=1.18 (0.96-1.44) for NHL, OR=1.47 (1.18-1.84) for DLCL, but not FL.	3586 NHL/ 4018	(37)
	IL1B	−31C>T	1143627	OR=0.73 (0.48-1.11) for NHL, 0.61 (0.36-1.04) for DLCL	241 NHL/372	(119)
	IL1RN	Intron 2, 86 bp VNTR		*2/*2 genotype: OR=5.51 (2.16-14.07) for GMZL	66 GMZL/163	(92)
	IL4	−1098T>G	2243248	OR=1.49 (1.05-2.11) for NHL, OR=3.84 (1.79-8.22) for T-NHL	518 NHL/597	(48)
				OR=1.59 (1.00-2.54) for DLCL	545 NHL/498	(50)
	IL4R	−29429C>T	2107356	OR =1.19 (0.91-1.55 for NHL, OR=2.6 (1.04-6.21) for T-NHL	518 NHL/597	(48)
				TT vs CC: OR=1.66 (1.13-2.42) for DLCL	1,172 NHL/982	(49)
	IL5	−745C>T	2069812	TT vs CC: OR=1.63 (1.04-2.54) for NHL, OR=2.28 (1.25-4.15) for DLCL	518 NHL/597	(48)
	IL6	−174 G>C	1800795	CC vs GG: OR=0.29 (0.1-0.87) for HL	88 young adult HL/their twin, 87	(54)
				NS for cHL, CC vs GG OR=0.13 (0.02-1.06) for NLPHL	408 HL/349	(55)
				OR=1.6 (0.85-3.01) for CTCL	63 CTCL/105	(120)
				NS for NHL, OR=0.46 (0.23-0.92) for T-NHL	518 NHL/597	(48)
		−596A>G,−597G>A,-	1800797	OR=2.68 (1.39-5.16) for CTCL	63 CTCL/105	(120)
				NS for NHL, OR=0.38 (0.19-0.78) for T-NHL	518 NHL/597	(48)
	IL7R	Ex4+33G>A	1494555	GG vs AA: OR=0.61 (0.38-0.99) for B-NHL; OR=2.32 (1.03-5.25) for T-NHL	518 NHL/597	(48)
	IL10	−819C>T	1800871	NS for NHL, DLCL, OR=1.48 (0.98-2.25) for FL	518 NHL/597	(48)
		−592C>A	1800872	CC vs AA: OR=1.6 (1.1-2.3) for AIDS related NHL	138 of AIDS ly,/1019 HIV+	(121)
				NS for NHL, DLCL, OR=1.47 (0.97-2.23) for FL	518 NHL/597	(48)
		−1082A>G	1800896	OR=1.97 (1.07-3.66) for aggressive lymphoma, HL: NS	156 NHL + HL	(122)
				G-allele frequency in DLCL (72%) vs. Controls (60%) p=0.023	199 DLCL/112	(123)
				OR=1.09 (0.97-1.22) for NHL, OR=1.17 (1.00-1.37) for DLCL	3586 NHL/ 4018	(37)
				GG vs AA: OR=1.53 (1.08-2.17) for NHL, OR=1.68 (1.15-2.44) for B-NHL	518 NHL/597	(48)
		−3575T>A	1800890	OR=1.11 (1.01-1.23) for NHL, OR=1.22 (1.06-1.40) for DLCL	3586 NHL/ 4018	(37)
				AA vs TT: OR=1.94 (1.31-2.87) for NHL, OR= 2.05 (1.36-3.11) for B-NHL	518 NHL/597	(48)
				AA vs TT: OR=1.84 (1.10-3.08) for DLCL	545 NHL/498	(50)
		Ex5+210T>C	3024496	CC vs TT: OR=1.55 (1.07-2.24) for NHL and OR=1.67 (1.13-2.47) for B-NHL	518 NHL/597	(48)
		IVS1-286G>T	3024491	TT vs GG: OR=1.54 (1.07-2.23) for NHL	518 NHL/597	(48)
	IL10RA	Ser159Gly A>G	3135932	OR=0.81 (0.65-1.02) for all lymphoma, OR=0.51 (0.31-0.84) for HL	678 NHL, HL, MM/669	(124)
	IL12B		3212227	OR=1.48 (1.03-2.12) for SLL	1,172 NHL/982	(49)
	IL13	Gln144Arg	20541	OR=1.53 (1.07-2.20) for SLL	1,172 NHL/982	(49)
		1069C>T	1800925	OR=1.50 (1.04-2.14) for SLL	1,172 NHL/982	(49)
	IFNGR1	IVS6-4G>A	3799488	NS for NHL, OR=0.52 (0.29-0.96) for FL	518 NHL/597	(48)
	IFNGR2	Ex7 −128C>T	1059293	OR=0.78 (0.59-1.03) for NHL, OR=0.67 (0.46-0.98) for DLCL	545 NHL/498	(50)
Chemokines	CCR5	delta32	333	OR=0.32 (0.13-0.79) for AIDS-related NHL	95 HIV-infected NHL/95 HIV+	(125)
				RR=0.20 (0-0.9) for AIDS related NHL	746 HIV-infected persons	(126)
	CXCL12	801G>A		3'A/+: OR=2.5 (1.3-4.8) for AIDS related NHL, 3'A/3'A: OR=5.1 (1.6-14)	746 HIV-infected persons	(126)
	IL8RB	Ex3-1010A>G	1126580	OR=0.77 (0.62-0.96) for NHL, GG vs AA: OR=0.45 (0.26-0.79) for DLCL	1,172 NHL/982	(49)
Innate Immunity	TLR2	–16933T>A	4696480	OR=2.82 (1.43-5.59) for FL, OR=0.61(0.38-0.95) for CLL	678 NHL, HL, MM/669	(124)
	TLR4	Asp299Gly	4986790	OR= 0.37 (0.13-1.03) for NHL	87 gastric MALT/594 HP+	(67)
				OR=0.92 (0.71-1.2) for NHL, OR=0.67 (0.45-0.99) for DLCL	904 NHL/1442	(64)
				OR=2.76 (1.12-6.81) for MALT, OR=1.80 (0.99-2.26) for HL	678 NHL, HL, MM/669	(124)
	TLR9	−1486T>C	187084	OR=1.61 (0.96-2.71) for FL	678 NHL, HL, MM/669	(124)
	CARD 15	1007fs	2066847	CC vs --: OR=3.1 (1.1-8.8) for NHL	904 NHL/1442	(64)
		Arg702Trp	2066844	OR=2.4 (1.0-5.6) for MALT lymphoma	83 gastric MALT/308 HP+	(65)
	FCGR2A	His165Arg	1801274	OR=1.31 (1.07-1.61) for NHL, effect seen for B- and T-cell lymphoma	1,172 NHL/982	(49)
				OR=1.50 (0.98-2.29) for DLCL	545 NHL/498	(50)
Oxidative Stress	NOS2A	Ser608Leu	2297518	Leu/Leu: OR=2.2 (1.1-4.4) for NHL, OR=3.4 (1.5-7.8) for DLCL and OR=2.6 (1.0-6.8) for FL	1,172 NHL/982	(72)
	MPO	−463 G>A	2333227	OR=1.3 (1.1-1.5) for NHL, OR=1.3 (1.0-1.7) for DLCL, OR=1.5 (1.0-2.4) for MZL	1,172 NHL/982	(72)
	SOD2	Val16Ala	4880	Ala/Ala: OR=1.3 (1.0-1.6) for B-cell Ly, OR=1.3 (1.0-1.89 for DLCL, NS for FL	1,172 NHL/982	(72)
				NS for NHL, OR=0.45 (0.60-1.12) for MZL	928 NHL/1446	(75)
	GPX1	Pro200Leu(Pro197Leu)	1050450	OR=1.31 (1.06-1.62) for NHL, effect seen in DLCL and FL	928 NHL/1446	(75)
	PPARG	His477His	3856806	OR=1.4 (1.0-2.1) for FL, OR=0.5 (0.3-0.9) for SLL, NS for DLCL	1,172 NHL/982	(72)
	GPX1	Pro200Leu(Pro197Leu)	1050450	OR=1.31 (1.06-1.62) for NHL, effect seen in DLCL and FL	928 NHL/1446	(75)
Energy Regulation
	LEP	−2548G>A	7799039	GA vs GG: OR=1.3 (1.0-1.7) for NHL; AA: OR=1.4 (1.0-1.9) for NHL	699 NHL/914	(82)
		19A>G	2167270	OR = 1.6 (1.1–2.3) for NHL, OR=1.9 (1.0-3.6) for FL	376 NHL/805	(81)
				19AA vs GG: OR=0.5 (0.3–0.8) for FL, NS for NHL and DLCL	699 NHL/914	(82)
	LEPR	Gln223Arg	1137101	NS for NHL, 223RR vs QQ: OR=1.9 (1.0–3.6) for FL among women	699 NHL/914	(82)
	NPY	−485T>C	16147	OR=1.5 (1.1-2.2) for NHL, OR=2.0 (1.2-3.6) for FL, NS for DLCL	308 NHL/684	(83)
		1258G>A	11557492	OR=1.6 (1.1-2.2) for NHL, OR=2.0 (1.1-3.5) for FL, NS for DLCL	308 NHL/684	(83)
		5671C>T	5574	OR=1.6 (1.1-2.2) for NHL, OR=1.8 (1.1-3.0) for FL, NS for DLCL	308 NHL/684	(83)
		1128 T>C	16139	OR=2.3 (1.1-4.9) for FL, NS for DLCL	308 NHL/684	(83)
	GHRL	−4427G>A	1629816	OR=0.78 (0.59-1.0) for NHL, OR=0.61 (0.40-0.94) for DLCL	308 NHL/684	(83)
Hormone Production
	CYP17A1	−34T>C	743572	CC vs TT: OR=1.4 (0.95-2.1) for NHL, OR=2.0 (1.1-3.5) for DLCL	308 NHL/684	(87)
				CC vs TT: OR=1.4 (1.02-2.03) for NHL, OR=1.76 (1.14-2.71) for DLCL	620 NHL/762	(88)
		137G>A	6162	AA vs GG: OR=1.4 (0.95-2.1) for NHL, OR=1.8 (1.0-3.3) for DLCL	308 NHL/684	(87)
		195C>A	6163	AA vs CC: OR=1.5 (0.99-2.2) for NHL, OR=2.1 (1.2-3.6) for DLCL	308 NHL/684	(87)
		−270A>C	3781287	CC vs AA: OR=1.5 (0.99-2.2) for NHL, OR=1.7 (0.92-3.0) for DLCL	308 NHL/684	(87)
		IVS2 105A>C	743575	CC vs AA: OR=1.5 (0.93-2.5) for NHL, OR=2.1 (1.1-4.2) for DLCL	308 NHL/684	(87)
				CC vs AA: OR=1.37 (0.91-2.05) for NHL, OR=1.61 (0.95-2.75) for FL	620 NHL/762	(88)
		75C>T	3740397	GG vs CC: OR=1.6 (1.1-2.5) for NHL, OR=2.3 (1.3-4.1) for DLCL	308 NHL/684	(87)
		2930G>T	4919685	TT vs GG: OR=2.2 (1.1-4.3) for DLCL	308 NHL/684	(87)
	COMT	108/158Val>Met	4680	NS for NHL, Met/Met: OR=2.0 (0.84-4.9) for FL in women	308 NHL/684	(87)
		701A>G	737865	NS for NHL, OR=0.42 (0.23-0.78) for FL in women	308 NHL/684	(87)
	PRL	−1149G>T	1341239	TT vs GG: OR=0.64 (0.41-1.0) for NHL, OR=0.53 (0.26-1.0) for FL	308 NHL/684	(87)
		−1488A>G	849877	GG vs AA: OR=0.70 (0.45-1.1) for NHL, OR=0.55 (0.27-1.1) for FL	308 NHL/684	(87)
Xenobiotic
	GSTT1	Null		OR=3.22 (1.94-3.22) for NHL	169 NHL/205	(91)
				OR=1.9 (1.04-3.46) for HL	90 HL/176	(93)
				OR=9.51 (4.57-19.81) for GMZL	66 GMZL/163	(92)
				OR=1.8 (1.1-3.0) for MALT	75 MALT/321	(94)
	GSTP1	Ile105Val (1548A>G)	947894	NS for NHL, Val/Val: OR=2.08 (1.05-4.14) for HL	219 HL and NHL/455	(95)
				Val/Val: OR=1.8 (0.6-5.1) for MZL, OR=0.2 (0.1-0.96) for DLCL	389 NHL/535	(96)
	PON1	Gln192Arg	662	Arg/Arg: OR=2.55 (1.37-4.55) for NHL	169 NHL/205	(91)
		Leu55Met	854560	Met/Met vs Leu/Leu : OR=1.36 (0.96-1.95) for NHL, effect seen for FL and T-NHL	1,172 NHL/982	(127)
	CYP1B1	Val432Leu	1056836	OR=1.27 (0.97-1.65) for NHL, effect seen across B-cell subtypes, but not T-cell lymphoma	1,172 NHL/982	(127)
	CYP2E1	Intron 6 DraI		CD: OR=0.62 (0.37-1.05) for NHL; CC: OR=0.68 (0.41-1.13) for NHL	219 HL, NHL/455	(95)
		−1054C>T (RsaI)	2031920	OR=0.59 (0.37-0.93) for NHL, effect seen in all histologic subtypes	1,172 NHL/982	(127)
	EPHX1	Tyr113His	1051740	Tyr/His: OR=0.45 (0.26-0.76) for lymphoma among males, His/His: NS	219 HL, NHL/455	(95)
		−1054C>T (RsaI)	2031920	OR=0.59 (0.37-0.93) for NHL, effect seen in all histologic subtypes	1,172 NHL/982	(127)
	NAT1	*3, *4, *10, *11		Slow acetylator genotype: OR=1.4 (0.9-2.3) for NHL in women, NS for men	389 NHL/535	(96)
		*3, *4, *10, *11, *14		*10*10 vs other genotypes: OR=1.60 (1.04-2.46) for NHL, OR=2.06 (1.12-3.79) for FL	1136 NHL/922	(101)
	NAT2	*4, *5, *6, *7		Intermediate/rapid vs slow: OR=1.21 (1.00-1.47) for NHL, OR=1.50 (1.11-2.02) for FL	1136 NHL/922	(101)
Cell Cycle Regulation
	CCND1	Pro241Pro	603965	AG vs GG: OR=1.1 (0.9-1.3) and AA vs GG: OR=1.4 (1.1-1.8) for NHL	1,172 NHL/982	(128)
Other
	BCL6	397G>C		OR=1.52 (1.13-2.05) for FL	85 FL/188	(129)
		–195C>T	1056932	CC vs TT: OR=2.2 (1.5-3.3) for NHL, effect seen in B-and T-NHL, CLL, DLCL, FL		(130)
	CASP3	Ex8-280C>A	6948	OR=0.7 (0.5-1.0) for B cell lymphoma, OR=0.5 (0.3-0.8) for FL	461 NHL/535	(131)
		Ex8+567T>C	1049216	OR=0.8 (0.6-1.0) for NHL, OR=0.4 (0.3-0.7) for FL	461 NHL/535	(131)
	CASP9	Ex5+32G>A	1052576	AA vs GG: OR=0.7 (0.5-1.0) for NHL, AA/AG vs GG: OR=0.6 (0.4-1.0) for FL	461 NHL/535	(131)
	CTLA4	−49A>G	231775	OR=2.0 (1.2-3.2) for NHL	100 NHL/128	(132)
				GG vs AA: OR=4.1 (0.9-18.2) for MALT lymphoma	62 MALT/250	(133)
		−318C>T	5742909	CT vs CC: OR=0.3 (0.1-0.9) for MALT lymphoma	62 MALT/250	(133)
		3′-UTR (AT)n		(AT)82 vs others: OR=1.6 (1.1-2.4) for NHL	100 NHL/128	(132)
	HSPA1B	Pst at 1267	1061581	P2P2: OR=18.2, p<=0.0001	44 NHL/106	(115)
	HSPA1L	Nco at 2437	2227956	OR=4.9 (p<=0.0002)	44 NHL/106	(115)
	JAK3	Ex24 291T>C	3008	OR=4.1 (1.23-13.67) for T-NHL	518 NHL/597	(48)
	NPAS2	Ala394Thr	2305160	OR=0.66 (0.51-0.85) for NHL, effect seen for DLCL and FL	455 NHL/527	(134)
	P73	Ex2 4G>A, 14C>T	2273953, 1801173	OR=1.46 (9.93-2.30) for NHL	103 NHL/440	(114)
	TGFB1	Leu10Pro	1982073	OR=0.71 (0.55-0.91) for NHL, OR=0.66 (0.47-0.93) for FL	545 NHL/498	(50)

Abbreviations: AIDS-Ly (AIDS-associated lymphoma), ATL (adult T-cell leukemia/lymphoma), BER (base excision repair); CTCL (cutaneous T-cell lymphoma), DLCL (diffuse large B-cell lymphoma), DSBR (double-strand break repair); GMALT (gastric mucosa-associated lymphoid tissue lymphoma), HP+ (Helicobacter pylori positive), MZL (marginal zone lymphoma), MR (mismatch repair); NHEJ (non-homologous end-joining); NS (not statistically significant), NER (nucleotide excision repair); PCNSL (primary central nervous system lymphoma), vs (versus).

Genetic polymorphisms that modify DNA integrity and methylation patterns influence lymphoma risk

SNPs in genes involved in DNA double-strand break and repair

The high risk of lymphoproliferative disorders in individuals carrying germline mutations in the ataxia telangiectasia, mutated (ATM) and Nijmegen breakage syndrome (NBS1) genes that involve syndromes associated with aberrant repair of DNA double-strand breaks underscores the relevance of this pathway in lymphomagenesis. Single nucleotide polymorphisms (SNPs) that hinder DNA repair mechanisms can increase the likelihood of pre-neoplastic lesions that may be relevant to lymphoma. For example, the WRN gene plays a crucial role in DNA double strand break repair and in other repair pathways. Mutations in this gene are associated with the autosomal recessive disorder, Werner syndrome, characterized by premature aging. More common genetic variants in WRN have been associated with NHL in two U.S. studies (12),(13). Specifically, the Arg and Ile alleles of two non-synonymous SNPs in the WRN gene (Cys1367Arg, Val114Ile) were underrepresented in NHL cases. Polymorphisms in five other non-homologous end joining /DNA double strand break repair genes (LIG4, RAG1, BRCA1, RCA2, XRCC3) and in the DNA double strand break repair checkpoint gene, TP53, also have been associated with NHL or specific subtypes (12),(13), (14). These studies highlight the relevance of genetic polymorphisms that may impede DNA double-strand break repair, important during normal B-cell maturation in the bone marrow or as a result of humoral immune response to antigen in germinal centers, in the risk of lymphoma.

Alkylating agents and other chemicals, oxidative stress, ionizing radiation and viruses provide other means to compromise genomic integrity. Mammalian cells have evolved numerous defense mechanisms to counter-balance these threats (15). Mismatch repair recognizes and repairs misplaced nucleotides and, if defective, can lead to microsatellite instability and neoplastic transformation. Two small studies report positive associations with a SNP in the mismatch repair gene, MSH2 (-6T>C), with NHL (16), (17) suggesting the potential relevance of this mechanism. However, there is limited and sometimes contradictory evidence for the role of base excision repair and nucleotide excision repair pathways and lymphoma risk (13),(12),(14).

Genetic polymorphisms in one-carbon metabolism and risk of lymphoma

Epigenetic silencing of tumor suppressor and B-cell specific genes is an important mechanism in lymphoma and in other hematopoietic malignancies (18;19). Genetic variants that influence methylation processes may promote lymphoma by mechanisms that involve hypo- or hypermethylation of proto-oncogenes or tumor suppressor genes, respectively, or through viral re-activation (20). Folate deficiency or genetic variation in folate metabolic pathways can influence DNA methylation patterns, as well as impede DNA synthesis and repair mechanisms. 5,10-methylenetetrahydrofolate reductase (MTHFR) is a key enzyme in the folate metabolic pathway that catalyzes the production of 5-methyltetrahydrofolate, a major donor of one-carbon units for DNA synthesis and methylation processes (Figure 1 online). A 677C>T SNP in the MTHFR gene, associated with enzyme thermolability (21), may hinder DNA methylation and shift the flux of one-carbon units toward purine and DNA synthesis and repair. Several NHL studies have examined this and other polymorphisms involved in folate metabolism, though results have been inconsistent (22), (23),(24), (25), (26), (27), (28), (29), (30)

Figure 1 online. Overview of the folate metabolic pathway.

S-adenosylmethionine (SAM); S-adenosylhomocysteine (SAH); tetrahydrofolate (THF); serine hydroxymethyltransferase (SHMT); 5,10-methylenetetrahydrofolate (methyleneTHF); 5,10-methylenetetrahydrofolate reductase (MTHFR); 5-methyltetrahydrofolate (methylTHF); methionine synthase (MTR); thymidylate synthase (TYMS); deoxythymidine monophosphate (dTMP); and deoxyuridine monophosphate (dUMP).

Similar discrepancies also have been reported for polymorphisms in the thymidylate synthase (TYMS) and methionine synthase (MTR) genes. TYMS plays a critical role in maintaining balanced supplies of deoxynucleotides required for DNA synthesis (Figure 1 online). Impairments of this enzyme have been associated with chromosome damage and fragile site induction (31), (32). A polymorphic 28-bp double (versus triple) repeat in the promoter region and a 6-bp deletion in the 3′UTR of the TYMS gene hinders TYMS gene expression and mRNA stability that may increase the rate of DNA double-strand breaks and chromosomal translocations. Positive associations with NHL have been reported for the 28-bp double repeat in an Asian study (33) and with the triple repeat in a British study (23), but no associations were found in a U.S. study (22). Furthermore, the 6-bp deletion has been positively (23) and inversely (22) associated with NHL. MTR catalyzes the re-methylation of homocysteine to methionine. The Gly allele of a SNP in the MTR gene (MTR 2756A>G, Asp919Gly), associated with lower homocysteine levels than the Asp allele, was inversely associated with lymphoma subtypes in Caucasians (28), (29), (30) and positively associated with NHL in another Caucasian study (22) and in an Asian population (34), (24).

Findings involving folate-metabolizing genes suggest a role for folate in lymphomagenesis that may involve DNA methylation and repair infidelity. The inconsistent findings in some studies may result from differences in folate status across populations since the influence of genetic variation on disease risk may be modified by folate status (35). Hence, studies are needed that consider the complex interrelationship between genetics, intakes of folate, vitamin B₆ and B₁₂, and other factors that may affect vitamin B status such as alcohol consumption.

SNPs that alter B-cell survival and growth influence lymphoma risk

Studies suggest that chronic inflammation may enhance lymphocyte neoplastic transformation by promoting proliferation and survival of mutated cells through activation of nuclear factor (NF)-κB and AP-1 response genes (36). Obesity, autoimmune disease, infection, chemical exposures and other inflammatory mediators that exacerbate oxidative stress can induce chronic inflammation. SNPs in genes involved in these processes may also play an important role in the pathogenesis of lymphoid neoplasms (Figure 2).

Figure 2. Nf-κB pathways and downstream factors affecting lymphoma risk.

Signaling through TNFRs, IL1R, TLRs and TCR leads to the activation of NF-κB. Genetic factors such as variants in TLR4, TNF, LEP and exogenous and endogenous factors such as microbial infections, oxidative stress and obesity may promote chronic inflammation and enhance cell survival by up-regulating proinflammatory cytokines such as TNF-α, IL-6, IL-1 and leptin. Activation of the NF-κB pathway can enhance transcription of anti-apoptotic mediators such as Bcl-2L1, Bcl-2A1, cIAPs, and SOD2 and proliferative factors such as cyclinD1, cyclinD2, and c-MYC (38). Promotion of cell proliferation and survival of mutated cells including those with aberrantly rearranged antigen-receptors may contribute to lymphomagenesis.

Tumor necrosis factor receptors (TNFRs), interleukin-1 receptor (IL1R), toll-like receptors (TLRs), T-cell receptor (TCR), nuclear factor kappa B (NF-κB), cyclooxygenase-2 (Cox-2), B-cell CLL/lymphoma 2-like 1 (Bcl-2L1), cellular Bcl2-related protein A1 (Bcl2A1), inhibitor-of-apoptosis proteins (cIAPs), superoxide dismutase-2 (SOD2), caspase recruitment domain family, member 15 (CARD15), nuclear factor kappa B inhibitor alpha (NFKBIA), interleukin-10 (IL10), IL-10 receptor alpha (IL10RA), tumor necrosis factor (TNF), leptin (LEP), interleukin-6 (IL6), cyclin-D1 (CCND1).

SNPs in pro-inflammatory cytokine genes influence risk of lymphoma

In perhaps the most noteworthy NHL risk alleles reported to date are SNPs in the tumor necrosis factor (TNF) and interleukin (IL)10 genes, reported by InterLymph, an international consortium of lymphoma epidemiological case-control studies (37). Researchers found that the TNF –308AA genotype conferred 25% and 65% increased risks for NHL and DLCL, respectively (37). Similarly, polymorphisms in the IL10 distal (-3575A>T) and proximal promoter (-1082A>G) regions, specifically the low IL-10 producer –3575A and –1082G alleles, were associated with modest increases in DLCL, but not FL risk. TNF-α, mainly produced by mast cells, macrophages and other immune cells, is a pro-inflammatory immunoregulatory cytokine and a key mediator of lymphocyte responses, natural killer cell activity and dendritic cell maturation. Elevated TNF-α expression augments anti-apoptotic behavior in B-cells through NF-κB activation that induces several anti-apoptotic factors including members of the Bcl2 family, cellular inhibitors of apoptosis, and cell cycle regulators (38) (Figure 2). The TNF – 308A allele has been associated with higher constitutional and inducible expression of TNF-α (39), (40) and increased susceptibility for rheumatoid arthritis and Sjogren syndrome (41), (42). IL-10 is a potent immunoregulatory cytokine of T-cells produced by monocytes and lymphocytes that hinders the inflammatory response by inducing apoptosis in mast cells and macrophages (43), thus inhibiting TNF-α production.(44-46). Genetic factors that up-regulate TNF-α or down-regulate IL-10 may provide a pro-inflammatory milieu that favors lymphomagenesis. Recent studies exploring other TNF and IL10 variants and haplotypes found further evidence for associations with NHL (47), (48), (49), (50). However, a number of studies reported conflicting results for another TNF variant (–857C>T) (50), (51), (52), (53). The TNF gene is located on chromosome 6p21 in the human leukocyte antigen (HLA) class III region in close proximity to several other immunoregulatory factors. Inconsistent results in variant-disease associations located in this region may be the result of population differences in linkage disequilibrium between the studied variant and the “causal” variant elsewhere in the HLA region. Fine mapping of the HLA region will help to clarify this and should be a major endeavor in future genetic studies of NHL.

IL-6 is a pro-inflammatory cytokine involved in regulation of immune defense mechanisms through initiation of acute phase responses. IL-6 serum levels have been positively associated with adiposity, type 2 diabetes and elevated levels have been detected in HL patients. Studies by InterLymph found no association between an IL6 promoter polymorphism (-174G>C) and NHL risk (37). However, inverse associations have been reported between the -174C allele and risk of young adult HL (54), young adult nodular lymphocyte predominant HL (55) and T-cell lymphoma (48), but no associations have been found for multiple myeloma or CLL (56), (57). Further, another IL6 variant (–598G>A) conferred a 60% decreased risk of T-cell lymphoma (58). Inverse associations have been reported between the IL6 –174C allele and systemic-onset juvenile chronic arthritis (59) and Kaposi sarcoma in HIV-positive men (60). An anti-inflammatory phenotype for the IL6 –174C allele has been recently described involving elevated apolipoprotein (Apo)-C1 levels and reduced heat shock protein 60 autoantibodies and platelet factor (PF)4 protein levels in serum of healthy adults (61). Inhibition of cholesterol ester transfer protein by Apo-C1 elevates the HDL/LDL ratio, an effect that promotes anti-oxidant/anti-inflammatory activities. Positive associations with PF4 plasma levels and hypercholesterolemia, inflammatory bowel disease and other pro-inflammatory conditions have been observed. Thus, the protective effects of the IL6 –174C allele for HL suggest the relevance of inflammatory and pro-oxidative processes in HL that will require further investigation.

SNPs in the innate immunity genes, toll-like receptor 4 (TLR4) and caspase recruitment domain family, member 15 (CARD15/NOD2) influence lymphoma risk

NOD2 and TLR4 are pro-inflammatory mediators integral as a first line of defense against viral and bacterial infection, providing non-specific protection against numerous pathogens. NOD2 orchestrates antimicrobial activity through a NF-κB-mediated pathway of inflammation and apoptosis (62). A rare CARD15 C insertion at nucleotide 1007 results in a premature stop codon that has been linked to autoimmune disorders such as Crohn's disease and psoriasis (63), and to an excess risk of lymphoma (64). Further, studies by the InterLymph consortium found that the CARD15 C insertion was associated with an elevated, but imprecise, risk estimate for NHL (OR = 2.3, 95% CI 0.12–135) (37). Yet another polymorphism in CARD15 (Arg702Trp) was recently associated with an elevated risk of MALT lymphoma among H. pylori infected individuals (65). Recent studies describe a CARD15 C insertion phenotype associated with reduced intestinal expression of α- and β-defensins and β-defensins (66) that may compromise the intestinal antimicrobial barrier. The link between Crohn's disease and NHL and the association of NHL and the CARD15 variants suggests that antigenic challenges to lymphoid tissue in the gastrointestinal tract can alter immune system responses and influence disease risk. Thus, further study of CARD15 mutations and the role of defensins in NHL may be warranted.

The TLR4 gene facilitates pro-inflammatory cytokine release by human cells in response to lipopolysaccharide, and mediates tolerance and B-cell activation. A TLR4 polymorphism (Asp299Gly) in the extracellular domain attenuates receptor signaling and reduces IL-12 and IFNγ levels. The TLR4 299Gly allele was inversely associated with risk of gastric MALT lymphoma (67) and DLCL (64), and positively associated with MALT lymphoma and HL (68). Nieters et al. also reported that a –16933T>A SNP in the TLR2 gene, another important pattern recognition receptor involved in the resolution of inflammation, increased risk of FL. These studies suggest that SNPs in TLR signaling pathways may differentially affect risk of lymphoma subtypes, but will require replication in larger study populations. Further, two studies reported that a functional variant (His165Arg) of an important innate immune response gene, the receptor for the F_C fragment of immunoglobulin G (FCGR2A), conferred an elevated NHL risk (49),(50). The Arg variant might favor lymphomagenesis based on impaired immunoglobulin G2-mediated phagocytosis (69) and promotion of antibody-based inflammation (70).

Oxidative stress genes are implicated in NHL

Reactive oxygen species (ROS) are implicated in several inflammatory conditions and in cancer risk. Phagocytic macrophages and neutrophils comprise the initial inflammatory response against infectious agents and antigens. These cells undergo a respiratory burst, where the membrane-bound enzymes, NADPH-oxidase and nitric oxide synthase (NOS), produce superoxide anion radicals and nitric oxide radicals. The Leu/Leu genotype of a non-synonymous SNP in the nitric oxide synthase gene (NOS2A Ser608Leu), located in a functionally relevant domain, has been associated with gastric cancer (71) and a 2-3-fold increased risk for NHL, DLCL and FL (72).(71). Bacterial infections such as H. pylori are characterized by extensive infiltration of neutrophils (73) that release myeloperoxidase (MPO), which amplifies the oxidative potential of hydrogen peroxides in affected tissues. Heterozygosity for a MPO -463G>A SNP was overrepresented among NHL cases, particularly for DLCL and MZL (72). Conversion of the superoxide anion to hydrogen peroxide is spontaneous or catalyzed by superoxide dismutase (SOD). Expression of the mitochondrial antioxidant enzyme, manganese SOD (SOD2), is induced by the pro-inflammatory cytokines, IL-1 and TNF-α. The variant Ala allele of a non-synonymous SNP in SOD2 (Val16Ala) that confers increased ROS scavenging (74) was associated with a marginal increased risk of B-cell lymphomas (72). However, a pooled analysis of 1,593 NHL cases and 2,517 controls found no association of this variant with overall NHL risk, but homozygosity for the 16Ala allele was associated with decreased MZL risk (75). In the same study, a variant in glutathione peroxidase 1 (GPX1 Pro197Leu) conferred 25% and 33% increased risks of NHL and FL, respectively (75). Glutathione peroxidase is one the most important antioxidant enzymes in humans and catalyzes the detoxification of hydrogen peroxide. These studies implicate pro-oxidant mechanisms that enhance free radical damage in increasing disease susceptibility. This may be particularly relevant in lymphocytes where chronic inflammation increases the risk of neoplastic changes such as is typified by the marginal-zone lymphomas.

SNPs in energy regulation genes are implicated in risk of NHL

An association between obesity and hemato-lymphopoietic cancers has become evident, with increased risks reported for NHL, leukemia, and multiple myeloma [reviewed in (76)]. Obesity is associated with impaired immune function (77) and generalized inflammation characterized by increased circulation of pro-inflammatory mediators such as leptin, TNF-α, IL-6, C-reactive protein and reduced levels of the anti-inflammatory peptide, ghrelin. Links between NHL and polymorphisms in the leptin (LEP), leptin receptor (LEPR), neuropeptide Y (NPY) and ghrelin (GHRL) genes may provide new insights in mechanisms of lymphomagenesis involving neuro-immune cross-talk. NPY acts on the central nervous system as a potent appetite stimulator controlled by positive and negative feedback actions of ghrelin and leptin, respectively, to regulate energy balance. Obesity upsets this fine balance, resulting in leptin resistance and elevated leptin and reduced ghrelin production. Leptin, NPY and ghrelin also exert major influences on humoral and cellular immune functions (78), (79), (80). Notably, leptin promotes proinflammory cytokine release, anti-apoptotic behavior in lymphocytes and oxidative stress (78). NPY suppresses innate immunity through natural killer cell activity inhibition, and increases adhesion, chemotaxis, phagocytosis, and superoxide anion production in macrophages. Ghrelin stimulates growth hormone release and inhibits proliferation of inflammatory cytokines such as IL-6, IL-1ß, and TNF-α.

In a U.S. based study, Skibola et al. found that the G allele for a SNP in the promoter region of the LEP gene (19A>G) conferred up to a two-fold increased risk of NHL, particularly FL (81). Further, gene-gene interaction was reported between LEP –2548G>A and LEPR Gln223Arg polymorphisms. Similar associations were observed in a U.K. study that found a 50% reduced FL risk associated with the LEP –19AA genotype and a 90% increased risk of FL among women associated with the LEPR 223ArgArg genotype (82). In the U.S. study population, four closely linked NPY variants were associated with up to a two-fold increased risk for NHL and FL (83). Of these, an NPY 1128T>C functional SNP may be particularly relevant as it has been associated with elevated NPY levels, enhanced angiogenesis, and lymphocyte proliferation (84). Furthermore, two variants in the GHRL gene (-4427G>A and 5179A>G) conferred 20-70% reduced risks for NHL, especially for DLCL, though no associations were found in the functionally relevant Leu72Met SNP (83) that has been associated with reduced BMI and percent abdominal visceral fat. These studies further implicate obesity in the pathogenesis of NHL that may involve its adverse action on immune system functions.

SNPs in genes involved in sex hormone production and metabolism influence lymphoma risk

Other less obvious but noteworthy loci linked to lymphoma are SNPs in genes that influence sex hormone production and metabolism. Prolactin and estrogens, important in female reproduction, also function as immune modulators that affect apoptosis, activation and proliferation of immune cells and promote B-cell proliferation and survival. Elevated prolactin levels have been associated with progression of hematological diseases such as multiple myeloma, AML and NHL (85;86). Two studies investigating SNPs in the CYP17A1 gene that encodes a key enzyme involved in estrogen and testosterone synthesis (87), (88) reported that the homozygous variant for a CYP17A1 34T>C SNP conferred a 40% increase in NHL risk in men and women, particularly for DLCL, where risks for the –34CC genotype were ~2-fold. The CYP17 –34CC genotype has been previously associated with elevated serum dehydroepiandrosterone sulphate and estradiol levels (89). Skibola et al. also found that SNPs in the catechol-O-methyltransferase (COMT) gene, involved in estrogen metabolism, and in the prolactin (PRL) gene, that influences lymphocyte prolactin levels, were associated with lymphoma risk in men and women (87). These results highlight a potential role of sex hormones in lymphoid tissue proliferation and ultimate neoplastic transformation. The positive association between obesity and lymphoma risk may reflect not only increases in pro-inflammatory mediators such as TNF-α, IL-6, leptin and others, but increased exposure to estradiol through aromatase-dependent conversion of androgens to estrogens in adipose tissue.

Limited evidence for detoxification genes in risk of lymphoma

Glutathione S-transferases (GSTM1, GSTT1, GSTP1) are involved in the detoxification of a wide range of carcinogens, including benzene, organochlorine compounds, organophosphate pesticides, tobacco smoke, chemotherapeutic agents, and reactive oxygen species (90). Deletion polymorphisms in GSTM1 (GSTM1*0) and GSTT1 (GSTT1*0) result in a loss of enzymatic activity. Evidence of an elevated risk for NHL and HL in GSTT1 null homozygotes was reported in several studies (91),(92), (93),(94). Proposed mechanisms could include an impaired neutralization of reactive oxygen species or reduced deactivation of carcinogenic intermediates of polycyclic aromatic hydrocarbons. No associations were found between GSTM1 and GSTP1 polymorphisms and NHL risk (95), (91), (96).

Homozygous variants of a paraoxonase-1 (PON1) Gln192Arg SNP that determines efficiency of hydrolysis and detoxification of specific organophosphates (97) was associated with a 2.5-fold elevated NHL risk (91). The 192Arg allele also has been associated with organophosphate toxicity in sheep farmers who dip animals in the pesticide diazinon (98), (99). Another study found no effect of the Gln192Arg variant, but an increased risk of FL and T-NHL associated with another non-synonymous variant, PON1 (Leu55Met) (100). Single reports also suggest some relevance of genetic variants in phase I cytochrome P450 enzymes (CYP1B1, CYP2E1) and epoxide hydrolase 1 (EPHX1) in the etiology of lymphomas (95), (100). These enzymes metabolize benzene, ethanol, halogenated solvents, and xenobiotic epoxide substrates, respectively. N-acetyltransferase (NAT) enzymes (NAT1, NAT2) catalyze the metabolization of aromatic and heterocyclic amines via N- or O-acetylation. One large study found a 60% increased risk associated with the NAT1*10/10 genotype and a 20% increased risk in intermediate and rapid NAT2 acetylators compared to slow acetylators (101). Interestingly, among intermediate/rapid acetylators, NHL risk was specifically elevated among current smokers compared with non-smokers (OR=2.44, 95%CI=1.15-5.20) (101) highlighting the importance of combined gene-environment analysis. Other smaller studies found only limited evidence for a role of variants in NAT-1 and -2 in lymphomagenesis (102), (91), (96).

Despite their crucial role in activation of potential carcinogens and detoxification of putative environmental lymphomagens, with the exception of GSTT1*0, genetic data on the contribution of xenobiotic metabolizing genes in the pathogenesis of lymphomas is scarce even in occupationally exposed populations.

Additional Studies Needed to Explore the HLA Region

The HLA region located on chromosome 6 (6p21.3) has now been mapped and approximately 220 genes have been defined. Many of these genes encode proteins involved in immune and inflammatory responses such as TNF, LTA, heat shock protein 70 and PRL (103). Few studies have explored the association of HLA genes with lymphoid malignancies; existing reports are mainly for HL. Two consecutive markers, D6S265 and D6S510, located in the HLA class I region have been identified as susceptibility loci for EBV-positive HL (104). Further, some studies have identified HL-associated HLA class II susceptibility alleles (105), (106), (107). What has emerged from these studies is that HLA-DPB1*0301 appears to confer susceptibility, and DPB1*0201 resistance to HL, and that HLA class II DRB1*1501 and DQB1*0602 alleles, or linked loci, may confer increased risk for sporadic and familial HL (105), (108). Thus far, few studies have explored the association of HLA polymorphisms with NHL risk and results have been inconclusive (109), (110). Fine map genotyping in the HLA region may help to clarify the significance of variability in this region for HL and other lymphoma subtypes.

Genome Screens

Few genome screens for lymphomas have been performed to date. Recently, a genome-wide linkage search of 115 families segregating CLL with or without additional B-cell lymphoproliferative disorders found evidence for a major susceptibility locus on the pericentric region of chromosome11 (chr11p11) (111). Linkage was suggested for four other chromosomal regions (5q22-23, 6p22, 10q25, and 14q32). Further, a genome screen on 44 high-risk families for HL identified strong linkage on chromosome 4 (near marker D4S394) and suggestive linkage on chromosomes 2 and 11 (112). Larger studies involving dense mapping in these regions and more exploratory studies involving whole genome scans for NHL will help to describe disease mechanisms and identify relevant pathways. To this end, plans for collaborative genome scan studies involving InterLymph consortium investigators currently are underway.

Conclusions

Case-control association studies provide further support for a genetic component to lymphoma. These studies suggest that disease mechanisms involving faulty DNA repair and genetic and environmental factors that deliver positive signals for B-cell survival and proliferation play an important role in lymphomagenesis (Figure 3). However, some genetic associations presented in this review may be false positive associations due to population stratification, improper control selection, genotyping error or other underlying causes. To date, most association studies have been limited by population size, restricting the potential investigation of gene-gene and gene-environment interactions. Larger collaborative studies using dense mapping to refine candidate regions and exploratory studies involving whole genome scans followed by replication and multi-center pooled analyses will help to eliminate false positive findings, further elucidate disease mechanisms and identify relevant pathways. InterLymph's successful identification of two susceptibility alleles in TNF and IL10 highlight the importance of consortia to validate biomarkers, particularly related to NHL subtypes. A thorough exposure assessment using biomarkers and the WHO-defined classification of lymphoma subtypes also will be needed to assess gene-environment interactions. Functional studies such as animal knockout or knockdown models and cellular models will help verify actual disease-predisposing variants to establish that the identified “causal variant” alters function and that the sequence change contributes to the illness phenotype. In summary, these studies will broaden our current understanding of important mechanistic pathways involved in lymphomagenesis and provide clues about environmental agents and lifestyle exposures that contribute to disease risk that may be translated to NHL screening, prevention or treatment regimens.

Figure 3. Environmental and genetic influences on fate of a naive mature B-cell encountering antigen.

In early and late stages of B-cell development, genetic polymorphisms and environmental exposures influence the fate of a B-cell and its chances of undergoing neoplastic transformation. Genetic polymorphisms that influence DNA repair (▲) such as XRCC3 and TYMS can increase the likelihood of chromosomal translocations and mutations that occur during normal B-cell maturation and as a result of genotoxic environmental exposures or endogenous genotoxic processes including class switch recombination and somatic cell hypermutation. Polymorphisms that impair DNA methylation (•) such as in the MTHFR and MTR genes may promote lymphoma by mechanisms involving hypo- or hypermethylation of proto-oncogenes or tumor suppressor genes. SNPs in genes that deliver positive signals for B-cell growth and survival (◆), such as in TNF, LEP and CYP17A1, or that block differentiation (■), such as in BCL-6, can enhance immortalization of B-cells. Oxidative stress genes such as superoxide dismutase (SOD2 and NOS2A) (★) may influence whether cells are protected from the harmful effects of reactive oxygen species.

Chronic antigenic stimulation of B-cells, through infection or proinflammatory conditions such as autoimmune disease or obesity, can activate B-cells and enhance their proliferation and survival. For lymphoma, this may be particularly relevant to growth, survival and ultimate transformation of B-cells that already carry pre-neoplastic lesions. SNPs can exacerbate these inflammatory responses (i.e., in pro-inflammatory cytokines, oxidative stress genes). The consistent associations found between B-cell NHL with genetic variants in pro-inflammatory factors such as TNF and leptin, and the association of viral, bacterial, and other exogenous agents leading to persistent inflammation, suggest this as one relevant mechanism underlying lymphomagenesis.