Welcome international healthcare professionals

This site is no longer supported and will not be updated with new content. You are welcome to browse and download all content already included in the site. Please note you will have to register your email address to access the site.

You are here

Cytokine gene polymorphisms and progression-free survival in classical Hodgkin lymphoma by EBV status: Results from two independent cohorts

Cytokine, 2, 64, pages 523 - 531

Highlights

 

  • Cytokines are important immune mediators in classical Hodgkin lymphoma.
  • SNPs in cytokine genes could influence cytokine production.
  • IL10, TNFA, IL6, IL1RN, INFG, CCL17 SNPs were not associated with cytokine levels.
  • Overall, none of these SNPs influence progression-free survival of CHL patients.
  • In analysis by EBV status, TNFA−308G>A influenced prognosis of EBV negative CHL.

Abstract

Background

Cytokines are important immune mediators of classical Hodgkin lymphoma (CHL) pathogenesis, and circulating levels at diagnosis may help predict prognosis. Germline single nucleotide polymorphisms (SNPs) in immune genes have been correlated with cytokine production and function.

Methods

We investigated whether selected germline SNPs in IL10 (rs1800890, rs1800896, rs1800871, rs1800872), TNFA (rs1800629), IL6 (rs1800795), ILRN (rs419598), INFG (rs2430561) and CCL17 (rs223828) were associated with circulating levels of related cytokines at diagnosis and progression-free survival (PFS) in CHL. Patients were from France (GELA, N = 464; median age = 32 years) and the United States (Iowa/Mayo Specialized Program Of Research Excellence [SPORE], N = 239; median age = 38 years); 22% of 346 CHL cases with EBV tumor status were positive.

Results

There was no association with any of the SNPs with cytokine levels. Overall, there was no association of any of the SNPs with PFS. In exploratory analyses by EBV status, TNFA rs1800629 (HRAA/AG = 2.41; 95%CI, 1.17–4.94) was associated with PFS in EBV-negative GELA patients, with similar trends in the SPORE patients (HRAA/AG = 1.63; 95%CI, 0.61–4.40). In a meta-analysis of the two studies, TNFA (HRAA/AG = 2.11; 95%CI, 1.18–3.77; P = 0.01) was statistically significant, and further adjustment for the international prognostic system did not alter this result.

Conclusions

This study showed that germline variation in TNFA was associated with CHL prognosis for EBV-negative patients, which will require confirmation. These results support broader studies on the differential impact of genetic variation in immune genes on EBV-positive vs. EBV-negative CHL pathogenesis.

Keywords: Hodgkin lymphoma, Cytokines, Polymorphism, TNFA, EBV.

1. Introduction

Classical Hodgkin lymphoma (CHL) is a highly curable lymphoid malignancy. However, approximately 15% of localized and 30% of disseminated disease patients do not respond or relapse after optimal initial therapeutic strategy and may require adapted first-line treatment. Several scoring systems using conventional biological and clinical parameters have been developed for identifying prognostic groups of patients in order to adapt the therapeutic strategy [1] . While scoring systems are useful in clinical practice, they do not fully reflect the biological heterogeneity of HL. Recent gene expression profiling of Hodgkin Reed-Sternberg (HRS) cells suggests distinct molecular profiles among CHL [2] . HRS cells create an immunologically favorable environment, and interaction between neoplastic cells and reactive cells is mediated by cytokines, chemokines and soluble receptors [3] .

The evaluation of cytokine levels at CHL diagnosis may provide prognostic information for patients [4] , and increased serum levels of sCD30 [5], [6], and [7], IL-10 [5], [6], [8], and [9], IL-6 [5], [10], and [11], TNFα [12] , IL-1Rα [5] , IL-2R [11] , BAFF [13] and CCL17 [14] and [15] have been correlated with patient outcome, but results are conflicting and limited by small studies, evaluation of single or only a small number of cytokines, lack of replication, and lack of prospective study designs [4] . In multivariate analysis integrating classical biological and clinical data, sCD30, IL-10, IL-6, IL-2R and IL-1Rα were each associated with outcome, with IL-6 showing the most consistent results [5] and [11]. More recently, the prognostic significance of the immune microenvironment has been confirmed, with data showing a correlation between an increased number of tumor-associated macrophages and a shortened survival in CHL patients [16] .

Germline polymorphisms have been described in immune genes, some of which appear to modify cytokine production in vitro and in vivo [17] . In the context of non-Hodgkin lymphoma (NHL), epidemiologic studies have suggested a role for germline single nucleotide polymorphisms (SNPs) in immune genes in disease risk, with the most robust (although not universal) association with the TNFA rs1800629 SNP [18], [19], [20], [21], and [22]. Clinical studies have also suggested that immune SNPs might also predict the disease prognosis. Two studies showed similar prognostic value for the TNFA SNP in diffuse large B-cell lymphoma (DLBCL) patients [23] and [24] but conflicting results were reported for IL10 promoter SNPs [24], [25], [26], [27], and [28]. The prognostic value of SNPs in IL4R, IL1A, IL8RB in DLBCL and IL8, IL2, IL12B, IL1RN in follicular lymphoma remained to be validated in independent series [24] and [29].

Relatively little is known about the role of cytokine gene polymorphisms in CHL. Five studies have explored the prognostic value of SNPs in IL6, TNFA and IL10 genes [28], [30], [31], [32], and [33], with conflicting results for IL6−174G>C SNP [30] and [31], consistent results for IL10−1082A>G and IL10−592>A SNPs [31] and [33], and no association for the IL10−3575T>A SNP [28] and [33] or the TNFA, IL13 and IL4R SNPs [31] and [33]. Herein, we investigated the association of cytokine gene polymorphisms in IL10, TNFA, IL6, IL1RN, INFG and CCL17 with cytokine levels and patient outcome in two independent cohorts of CHL patients. Recent epidemiological studies suggest that genetic risk factors differ between EBV+ and EBV CHL [34] and gene expression profiling studies have demonstrated that EBV status of the CHL tumor may delineate distinct biological entities with specific microenvironment and immune regulation [35] . Large studies have confirmed differences in clinical characteristics between EBV+ and EBV CHL, with a poorer prognosis for EBV+ CHL patients older than 50 years old [36] . In this context, we analyzed the correlation between candidate immune SNPs and disease progression by EBV status, which to our knowledge has not been previously reported.

2. Materials and methods

2.1. Study populations

The first cohort consisted of 464 CHL patients enrolled between 1998 and 2002 and followed through 2009 as part of a prospective study of the GELA assessing the prognostic values of plasma cytokines and soluble receptors [5] . In accordance with French law, race or ethnicity could not be collected; however, given the national scope of this GELA study, the cohort likely reflects the genetic heterogeneity of the French population, which is dominated by Western-European ancestry, as shown by principal components analysis from genome-wide association studies in this population [37] . The second cohort consisted of 239 newly diagnosed CHL patients age 18 years and older prospectively enrolled from 2002 to 2009 and followed through 2011 in the University of Iowa/Mayo Clinic SPORE (Specialized Program of Research Excellence) Molecular Epidemiology Resource [38] . In the SPORE cohort, 202 patients were Caucasian, and the remainder were non-Caucasian (n = 6) or unknown/refused (n = 31). Patients in both cohorts were negative for human immunodeficiency virus. Both cohorts were systematically followed for disease progression, retreatment and death. A peripheral blood sample for serum, plasma and DNA were collected from all patients at diagnosis in the two studies, and DNA for genotyping was extracted using standard protocols. These studies were approved by the ethics committees of Dijon and Lyon University Hospitals for the GELA study and by the Human Subjects Institutional Review Board at Mayo Clinic and the University of Iowa for the SPORE study, and all patients provided written consent for participation.

2.2. Laboratory analysis

2.2.1. Genotyping

SNPs from candidate genes included IL10 (rs1800890, rs1800896, rs1800871, rs1800872), TNFA (rs1800629), IL6 (rs1800795) and CCL17 (rs223828), which were all located in the promoter region; SNPs in ILRN (rs419598) and INFG (rs2430561) were localized in exon 5 (synonymous SNP) and intron 1, respectively. These SNPs were chosen based on i) their localization in regulatory regions of the genes (IL10, TNFA, IL6, CCL17); ii) previous studies showing in vitro functional effect of these SNPs on cytokine expression [17] and [39]; iii) previous lymphoma studies showing that these SNPs might predict clinical outcome [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], and [33].

In GELA, genotyping used specific fluorescent dye-labeled (FAM and VIC) MGB probes (Applied Biosystems, Foster City, California, USA). Real-time PCR analysis was performed on an ABI Prism 7000 Sequence Detection System (Applied Biosystems). All genotyping was performed in duplicate, and agreement was 100%. In the SPORE, the candidate SNPs were genotyped as part of a larger project using a custom Illumina Infinium array (Illumina, San Diego, CA). Standard genotyping quality control procedures were performed and included duplicate genotyping, dropping samples or SNPs with call rates <95%, and testing for Hardy–Weinberg equilibrium (HWE). We found >99.9% genotyping concordance among the 3502 samples with duplicates.

2.2.2. Measurement of cytokines

In GELA, data on plasma cytokine levels were available for IL-10, TNFα, IL-6 and IL1-Rα, as previously described [5] . Briefly, plasma samples collected before any treatment (including corticosteroid administration) and were immediately stored at −80 °C in sterile EDTA-containing tubes. Levels were measured in a centralized referent laboratory in duplicate using ELISA kits obtained from Biosource (Nivelles, Belgium) for TNFα, R&D system (Minneapolis, NM, USA) for IL-1Rα and Beckman-Coulter (Fullerton, CA, USA) for IL-6 and IL-10. The detection limit of the test was 5 pg/ml for IL-10, 3 pg/ml for TNFα, 3 pg/ml for IL-6 and 22 pg/ml for IL-1Rα.

For the SPORE cohort, serum cytokines were measured from pre-treatment blood draws using a multiplex ELISA (Invitrogen, Camarillo, CA, USA). Thirty serum cytokines, including data on IL-10, TNFα, IL-6, IL-1Rα, CCL17 reported here, were analyzed using the Luminex-200 system Version 1.7 (Luminex, Austin, Texas, USA) [11] . Data were acquired using STar Station software (Applied Cytometry, Dinnington, Sheffield, UK) and analysis was performed using the MasterPlex QT 1.0 system (MiraiBio, Alameda, CA). Cytokines measured below the limit of detection were assigned a value of (limit of detection/2). Inter-assay variation was assessed by inclusion of an internal control serum on all assay plates; the CV for the internal control log2 cytokine values across plates ranged from 0.3% (RANTES) to 12.2% (IL-6).

2.2.3. Epstein Barr Virus (EBV) status of the tumor

Tumor samples were available for 242 of 464 GELA CHL patients (52%), and EBV status was determined by immunohistochemical detection of latent membrane protein-1 (LMP-1) in 186 patients (77%), by EBER in situ hybridization in 48 patients (20%) or by other techniques in the remaining 8 patients (3%). Tumor samples were available 104 of the 239 SPORE CHL (48%), and EBV status was determined by EBER in situ hybridization.

2.3. Statistical analysis

Progression free survival (PFS) was measured from the date of initiation of therapy (GELA series) or the date of diagnosis (SPORE series) to the date of first relapse, progression, retreatment, or death (any cause); patients without an event were censored at date of last follow-up. Associations between cytokines and genotype were assessed using the Kruskal–Wallis test. Kaplan–Meier curves and Cox proportional hazards models were used to assess the association of each genotype with PFS in both a univariate fashion and after adjusting for the international prognostic score (IPS). A dominant genetic model for each candidate SNP was used based prior publications. Analyses were done independently for the two cohorts and then to increase power, a meta-analysis approach was used to combine results [40] . Analyses were performed using SAS version 9.2 (SAS Institute Inc., Cary, NC, USA) and R ( http://www.r-project.org/ ).

3. Results

3.1. Patient characteristics and genotyping of immune SNPs

Clinical characteristics of the patients from the two cohorts are demonstrated in Table 1 . In the GELA cohort, the median age was 32 years (range, 15–93), 57% were male, 44% had B-symptoms at diagnosis, 72% had an Ann Arbor stage I–II, and 73% had an IPS score of 0–2. Nodular sclerosis was the main CHL histologic subtype (384 patients, 83%). The EBV status of the tumor was established for 242/464 patients (52%), and was positive in 53 (22%) and negative for 189 (78%) patients. A total of 277 (60%) patients were treated with ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine), 102 (22%) with EBVP (epirubicin, bleomycin, cyclophosphamide, vincristine, prednisone), 60 (13%) with BEACOPP (bleomycin, etoposide, doxorubin, cyclophosphamide, vincristine, procarbazine, prednisone) and 25 (5%) with other anthracycline-based regimens.

Table 1 Clinical and biologic characteristics of classical Hodgkin lymphoma patients.

  GELA cohort SPORE cohort
  All EBV+ EBV P a All EBV+ EBV P a
  N = 464 N = 53 N = 189   N = 239 N = 23 N = 81  
  N (%) N (%) N (%)   N (%) N (%) N (%)  
Median age 32 37 33 0.127 38 27 38 0.26
 Range 15–93 17–89 15–93   18–89 21–73 18–81  
Age ⩾45 years 101 (22) 13 (25) 46 (24) 0.98 92 (39) 5 (22) 30 (37) 0.17
Male 264 (57) 43 (81) 103 (54) 0.001 124 (52) 17 (74) 40 (49) 0.09
ECOG PS                
 0–1 395 (97) 42 (98) 158 (95) 0.40 205 (86) 17 (74) 68 (84) 0.36
 2–4 12 (3) 1 (2) 9 (5)   34 (14) 6 (26) 13 (16)  
 
Histological subtype
 Nodular sclerosis 384 (83) 39 (74) 161 (85) 0.07 b 161 (67) 14 (61) 62 (77) 0.06 b
 Mixed cellularity 40 (9) 7 (13) 11 (6)   27 (11) 5 (22) 7 (9)  
 Diffuse lymphocyte predominance 5 (1) 0 (0) 2 (1)   12 (5) 1 (4) 6 (7)  
 Lymphocyte depletion 2 (<1) 1 (2) 0 (0)   0 (0) 0 (0) 0 (0)  
 Unclassified 33 (7) 6 (11) 15 (8)   39 (17) 3 (13) 6 (7)  
 
Ann Arbor stage
 I–II 335 (72) 29 (55) 105 (56) 0.76 122 (52) 10 (45) 39 (49) 0.79
 III–IV 129 (27) 24 (45) 84 (44)   113 (48) 12 (55) 41 (51)  
B symptoms 206 (44) 28 (53) 91 (48) 0.55 92 (39) 14 (61) 27 (33) 0.02
Hemoglobin level <10.5 g/dl 64 (14) 11 (21) 30 (16) 0.62 32 (14) 3 (14) 11 (14) 0.97
ESR ⩾ 50 196 (44) 17 (34) 91 (51) 0.04 50 (34) 6 (43) 18 (38) 0.73
White-cell count ⩾15,000/mm3 85 (18) 4 (8) 41 (22) 0.01 30 (13) 2 (10) 15 (19) 0.31
Lymphocyte count <600/mm3 40 (9) 3 (6) 21 (11) 0.20 29 (13) 6 (29) 12 (15) 0.13
Albumin level <40 g/l 191 (45) 22 (42) 88 (47) 0.15 103 (51) 9 (45) 34 (48) 0.82
 
Hasenclever score
 0 60 (14) 3 (6) 21 (12) 0.94 0 (0) 0 (0) 0 (0) 0.92
 1 139 (33) 16 (33) 47 (28)   36 (15) 3 (13) 12 (15)  
 2 111 (26) 13 (27) 37 (22)   84 (35) 10 (44) 26 (32)  
 3 64 (15) 11 (22) 33 (20)   64 (27) 3 (13) 23 (28)  
 ⩾4 47 (12) 6 (12) 29 (18)   55 (23) 7 (30) 20 (25)  

a P-value compares EBV+ and EBV classical Hodgkin lymphoma.

b P-value compares nodular sclerosis and mixed cellularity histology subtype.

EBV, Epstein Barr Virus; ECOG, Eastern Cooperative Oncology Group; PS, performance status.

The SPORE cohort was similar to the GELA cohort. The median age was 38 years (range, 18–89), 52% were male, 39% had B-symptoms at diagnosis, 52% had an Ann Arbor stage I–II, and 50% had an IPS score of 0–2. Nodular sclerosis was also the most common subtype (161 patients, 67%). The EBV status of the tumor was established for 104/239 patients (48%), and was positive for 23 (22%) and negative for 81 (78%) patients. ABVD regimen was used in 214 (90%) patients, 12 (5%) patients received Stanford V protocol, three patients (1%) were treated by MOPP (caryolysine, vincristine, vinblastine, procarbazine) and 10 (4%) with other regimens.

Details of the genotype distributions are presented in Table 2 . The minor allele frequencies (MAF) of the 9 immune SNPs were very similar between the two series, with the exception of TNFA, which had an MAF of 0.12 in GELA compared to 0.20 in the SPORE. Genotype distributions for each SNP by EBV status were consistent with the overall distributions in each cohort.

Table 2 Genotype distribution of immune SNPs in Hodgkin lymphoma, overall and by EBV status.

  Genotype GELA cohort SPORE cohort
    All EBV+ EBV P All EBV+ EBV P
    N (%) N (%) N (%)   N (%) N (%) N (%)  
IL10−3575T>A N 448 49 182   239 23 81  
rs1800890 TT 200 (44) 22 (45) 86 (47) 0.95 112 (47) 13 (57) 36 (45) 0.24
  TA 204 (46) 23 (47) 81 (45)   97 (41) 9 (39) 35 (43)  
  AA 44 (10) 4 (8) 15 (8)   30 (12) 1 (4) 10 (12)  
 
IL10−1082A>G N 459 52 186   239 23 81  
rs1800896 AA 154 (34) 17 (33) 69 (37) 0.84 82 (34) 11 (48) 27 (33) 0.35
  AG 232 (50) 27 (52) 91 (49)   107 (45) 8 (35) 40 (50)  
  GG 73 (16) 8 (15) 26 (14)   50 (21) 4 (17) 14 (17)  
 
IL10−819C>T N 446 49 181   239 23 81  
rs1800871 CC 251 (56) 26 (53) 99 (55) 0.76 128 (54) 10 (43) 42 (52) 0.50
  CT 165 (37) 20 (41) 66 (36)   99 (41) 11 (48) 33 (41)  
  TT 30 (7) 3 (6) 16 (9)   12 (5) 2 (9) 6 (7)  
 
IL10−592C>A N 447 49 182   239 23 81  
rs1800872 CC 251 (56) 26 (53) 99 (54) 0.77 128 (54) 10 (43) 42 (52) 0.50
  AC 166 (37) 20 (41) 67 (37)   99 (41) 11 (48) 33 (41)  
  AA 30 (7) 3 (6) 16 (9)   12 (5) 2 (9) 6 (7)  
 
TNFA−308G>A N 464 53 189   239 23 81  
rs1800629 GG 357 (77) 41 (77) 148 (78) 0.61 154 (65) 14 (61) 54 (67) 0.58
  AG 98 (21) 12 (23) 38 (20)   75 (31) 8 (35) 25 (31)  
  AA 9 (2) 0 (0) 3 (2)   10 (4) 1 (4) 2 (2)  
 
IL6−174G>C N 201 24 84   239 23 81  
rs1800795 GG 85 (42) 10 (42) 35 (42) 0.56 84 (35) 10 (43) 24 (29) 0.21
  CG 85 (42) 12 (50) 35 (42)   112 (47) 10 (43) 41 (51)  
  CC 31 (16) 2 (8) 14 (16)   43 (18) 3 (13) 16 (20)  
 
IL1RN+2018T>C N 199 23 83   239 23 81  
rs419598 TT 109 (55) 13 (57) 45 (54) 0.88 132 (55) 17 (74) 44 (54) 0.13
  TC 78 (39) 9 (39) 32 (39)   88 (37) 4 (17) 29 (36)  
  CC 12 (6) 1 (4) 6 (7)   19 (8) 2 (9) 8 (10)  
 
INFG+874A>T N 200 24 83   239 23 81  
rs2430561 AA 54 (27) 7 (30) 24 (29) 0.89 61 (26) 6 (26) 20 (25) 0.58
  TA 97 (48) 12 (50) 38 (46)   129 (54) 15 (65) 48 (59)  
  TT 49 (25) 5 (20) 21 (25)   49 (20) 2 (9) 13 (16)  
 
CCL17−431C>T N 198 24 82   239 23 81  
rs223828 CC 178 (90) 22 (92) 75 (91) 0.97 213 (89) 20 (87) 74 (91) 0.53
  TC 19 (10) 2 (8) 7 (9)   25 (11) 3 (13) 7 (9)  
  TT 1 (<1) 0 (0) 0 (0)   1 (<1) 0 (0) 0 (0)  

3.2. Correlation between genotypes and cytokine levels

Based on the results of the initial prospective studies of the prognostic value of cytokine levels [5] and [11], we evaluated the correlation between immune SNPs and plasma cytokine levels in the GELA cohort and serum cytokine levels in the SPORE cohort. IL-10, TNFα, IL-6 and IL-1Rα were evaluated in both cohorts, while CCL17 was only evaluated in the SPORE. In GELA, there was no association of plasma cytokine levels and their respective genotypes for IL10 (4 SNPs), TNFA, IL6 and IL1RN ( Table 3 ). In the SPORE cohort, there was also no association with serum cytokine for IL10, TNFA, IL6, IL1RN and their respective genotypes. Nor was there any association of CCL17 genotypes with the serum level of this chemokine ( Table 4 ). When we compared cytokine levels between EBV+ and EBV CHL, the median level of TNFα (36 vs. 27.5 pg/ml, P = 0.005) was higher in EBV+ compared to EBV CHL in the GELA. The finding for TNFα (serum) was not observed in the SPORE cohort. For other cytokines, we did not observe any difference of the median level between EBV+ and EBV CHL in the GELA or SPORE cohorts (data not shown).

Table 3 Plasma cytokine levels by SNP genotype, GELA cohort.

    Whole cohort EBV+ EBV
N (%) Median Range P N (%) Median Range P N (%) Median Range P
IL10−3575T>A                          
rs1800890 TT 196 (44) 12 0–2952 0.29 20 (43) 17 0–326 0.13 85 (47) 12 0–2952 0.42
n = 442 TA 202 (46) 11 0–24   23 (49) 9 0–96   79 (45) 7 0–194  
  AA 44 (10) 6 0–137   4 (8) 0 0–17   15 (8) 12 0–86  
 
IL10−1082A>G                          
rs1800896 AA 151 (33) 11 0–2952 0.76 16 (33) 25.5 0–316 0.08 68 (37) 8 0–2952 0.87
n = 452 AG 229 (51) 11 0–932   26 (53) 16.1 0–96   89 (49) 10 0–194  
  GG 72 (16) 9 0–166   7 (14) 0 0–32   26 (14) 11.3 0–108  
 
IL10−819C>T                          
rs1800871 CC 249 (57) 11 0–931 0.80 25 (53) 17 0–147 0.39 98 (56) 11 0–113 0.43
n = 440 CT 161 (36) 9 0–2952   19 (40) 6 0–316   65 (37) 7 0–2952  
  TT 30 (7) 14 0–183   3 (7) 15 0–41   16 (7) 24 0–183  
 
IL10−592C>A                          
rs1800872 CC 249 (56) 11 0–931 0.80 25 (53) 17 0–147 0.39 98 (56) 11 0–113 0.45
n = 440 AC 162 (37) 9 0–2952   19 (40) 6 0–316   65 (37) 7 0–2952  
  AA 30 (7) 14 0–183   3 (7) 15 0–41   16 (7) 24 0–183  
 
TNFA−308G>A                          
rs1800629 GG 357 (77) 27 6–248 0.17 38 (76) 37 14–143 0.48 144 (78) 28 6–248 0.73
n = 464 AG 98 (21) 26 8–236   12 (24) 33 12–147   37 (20) 27 13–133  
  AA 9 (2) 21 11–36   0 (0)   3 (2) 21 16–36  
 
IL6−174G>C                          
rs1800795 GG 84 (43) 15 0–390 0.33 9 (43) 24 0–75 0.21 35 (42) 15 0–390 0.68
n = 197 CG 84 (43) 15 0–99   11 (52) 16 4–99   35 (42) 12 0–66  
  CC 29 (14) 22 2–300   1 (5) 300   13 (16) 20 3–300  
 
IL1RN+2018T>C                          
rs419598 TT 108 (55) 500 20–5000 0.85 12 (55) 389 82–4614 0.70 45 (55) 559 141–3500 0.59
n = 197 TC 77 (39) 540 26–4000   9 (41) 700 168–1650   31 (38) 531 189–1400  
  CC 12 (6) 523 242–1300   1 (4) 423   6 (7) 740 367–1300  

Table 4 Serum cytokine levels by SNP genotype, SPORE cohort.

    Whole cohort EBV+ EBV
N (%) Median Range P N (%) Median Range P N (%) Median Range P
IL10−3575T>A                          
rs1800890 TT 59 (46) 23 3–1352 0.64 7 (50) 22 10–59 0.12 19 (42) 18 3–1106 0.22
n = 129 TA 55 (43) 27 3–167   7 (50) 35 11–167   20 (45) 26 2–80  
  AA 15 (11) 28 6–89   4 (0)   6 (13) 30 22–36  
 
IL10−1082A>G                          
rs1800896 AA 42 (33) 22 3–1106 0.40 6 (43) 22 10–126 0.47 12 (27) 19 3–1106 0.31
n = 129 AG 58 (45) 27 3–1352   7 (50) 27 11–167   25 (55) 26 3–80  
  GG 29 (22) 28 3–89   1 (7) 65   8 (18) 30 15–37  
 
IL10−819C>T                          
rs1800871 CC 71 (55) 27 3–1352 0.57 5 (36) 27 16–65 0.39 24 (53) 27 3–517 0.67
n = 129 CT 51 (40) 24 3–1106   7 (50) 29 10–107   18 (40) 19 3–1106  
  TT 7 (5) 24 8–61   2 (14) 35 10–59   3 (7) 23 8–37  
 
IL10−592C>A                          
rs1800872 CC 71 (55) 27 3–1352 0.57 5 (36) 27 16–65 0.39 24 (53) 27 3–517 0.67
n = 129 AC 51 (40) 24 3–1106   7 (50) 29 10–107   18 (40) 19 3–1106  
  AA 7 (5) 24 8–61   2 (14) 35 10–59   3 (7) 23 8–37  
 
TNFA−308G>A                          
rs1800629 GG 81 (63) 6 3–333 0.32 10 (71) 3 3–333 0.48 28 (62) 3 3–242 0.59
n = 129 AG 44 (34) 3 3–44   4 (29) 3 3–34   16 (36) 4 3–10  
  AA 4 (3) 5 3–231   0 (0)   1 (2) 3  
 
IL6−174G>C                          
rs1800795 GG 45 (35) 12 2–409 0.87 7 (50) 39 6–130 0.21 15 (34) 10 2–409 0.94
n = 129 CG 68 (53) 13 2–1430   6 (43) 6 1–63   24 (53) 12 1–1430  
  CC 16 (12) 7 2–252   1 (7) 81   6 (13) 10 2–70  
 
IL1RN+2018T>C                          
rs419598 TT 68 (53) 644 15–4459 0.33 10 (71) 1176 15–4459 0.20 22 (49) 587 331–1182 0.07
n = 129 TC 52 (40) 676 64–4762   4 (29) 606 394–1042   19 (42) 734 516–3189  
  CC 9 (7) 550 184–1225   0 (0)   4 (9) 685 184–1225  
 
CCL17−431C>T                          
rs223828 CC 101 (91) 1633 107–4832 0.53 10 (83) 1015 181–4004 0.13 37 (95) 2170 239–4832 0.44
n = 111 TC 10 (9) 1767 325–3452   2 (17) 2526 1894–3158   2 (5) 1333 511–2155  
  TT 0 (0)   0 (0)   0 (0)  

3.3. Immune SNPs and outcome

After a median follow-up of 6.3 years (range, 0.3–8.2), the 6-year PFS was 81.5% (95% confidence interval [CI], 74.1–84.1) with 78 (17%) patients with a treatment failure (progression or relapse) in the GELA cohort. After a median follow-up of 5 years (range 0.04–9.75), the 6-year PFS was 72.9% (95% CI, 64.8–79.7) with 57 (24%) patients with a treatment failure in the SPORE cohort. The 6-year PFS was similar for EBV+ and EBV CHL (78.3% [68–91] vs. 82.5% [77–88], P = 0.25) in the GELA cohort, but EBV patients had a higher PFS rate compared to EBV+ CHL in the SPORE cohort (71.9% [58–89] vs. 29.0% [7–100], P = 0.007) noting that EBV+ CHL was based on only 23 cases and 11 events.

In both studies for all CHL, there was no association of any of the nine SNPs with PFS ( Table 5 ). The association of SNPs with PFS was subsequently analyzed by tumor EBV status ( Table 6 ). In EBV+ CHL, there were no associations of any SNPs with PFS in the GELA cohort, while in the SPORE cohort, there was an association of the CCL17 with PFS (hazard ratio [HR]=6.97, 95%CI 1.54–31.46; P = 0.01), but this was based on a small number of events. In EBV CHL, TNFA−308G>A was associated with PFS in the GELA cohort (HR = 2.41, 95%CI 1.17–4.94; P = 0.02). In the SPORE cohort a trend in the same direction was observed (HR = 1.63, 95%CI 0.61–4.40), although this was not statistically significant (P = 0.33). Survival curves comparing PFS of EBV CHL patients with TNFA−308GG vs. AG + AA genotypes are shown in Fig. 1 . In a meta-analysis of the two studies, the pooled estimate was statistically significant (HR = 2.11, 95%CI 1.18–3.77; P = 0.01), and further adjustment for IPS strengthened this result (HR = 2.52, 95%CI 1.39–4.58; P = 0.002).

Table 5 Association of immune SNPs with progression-free survival in classical Hodgkin lymphoma.

Gene SNP Model a GELA cohort SPORE cohort Combined
MAF HR (95%CI) P MAF HR (95%CI) P HR (95%CI) P
IL10−3575T>A rs1800890 TT vs. TA + AA 0.33 1.04 (0.65, 1.64) 0.87 0.33 0.83 (0.49, 1.39) 0.47 0.94 (0.67, 1.33) 0.73
IL10−1082A>G rs1800896 AA vs. AG + GG 0.41 1.07 (0.66, 1.73) 0.77 0.43 1.20 (0.68, 2.11) 0.52 1.13 (0.78, 1.62) 0.52
IL10−819C>T rs1800871 CC vs. CT + TT 0.25 1.34 (0.85, 2.10) 0.21 0.25 0.82 (0.49, 1.39) 0.47 1.09 (0.77, 1.53) 0.63
IL10−592C>A rs1800872 CC vs. AC + AA 0.25 1.33 (0.84, 2.09) 0.22 0.25 0.82 (0.49, 1.39) 0.47 1.08 (0.77, 1.53) 0.64
TNFA−308G>A rs1800629 GG vs. AG + AA 0.13 1.15 (0.69, 1.91) 0.58 0.20 1.19 (0.70, 2.03) 0.52 1.17 (0.81, 1.69) 0.40
IL6−174G>C rs1800795 GG vs. GC + CC 0.37 0.73 (0.39, 1.37) 0.33 0.41 1.30 (0.74, 2.30) 0.36 1.00 (0.66, 1.53) 0.98
IL1RN+2018T>C rs419598 TT vs. TC + CC 0.20 0.80 (0.41, 1.54) 0.51 0.26 0.86 (0.51, 1.45) 0.57 0.81 (0.53, 1.23) 0.32
IFNG+874A>T rs2430561 AA vs. TA + TT 0.51 0.81 (0.40, 1.62) 0.55 0.47 0.78 (0.44, 1.38) 0.39 0.79 (0.51, 1.23) 0.30
CCL17−431C>T rs223828 CC vs. TC + TT 0.05 1.06 (0.37, 2.96) 0.92 0.06 0.93 (0.41, 2.09) 0.85 0.97 (0.51, 1.84) 0.93

a Dominant model.

MAF, minor allele frequency; HR, hazard ratio; CI, confidence interval.

Table 6 Association of immune SNPs with progression-free survival in classical Hodgkin lymphoma by EBV status.

SNP a EBV negative CHL EBV positive CHL
GELA cohort SPORE cohort Combined GELA cohort SPORE cohort Combined
HR (95%CI) P HR (95%CI) P HR (95%CI) P HR (95%CI) P HR (95%CI) P HR (95%CI) P
IL10−3575T>A 1.20 (0.59, 2.42) 0.60 0.63 (0.23, 1.69) 0.36 0.97 (0.55, 1.72) 0.91 0.69 (0.19, 2.38) 0.73 1.01 (0.27, 3.76) 0.99 0.82 (0.33, 2.03) 0.67
IL10−1082A>G 1.39 (0.65, 2.92) 0.39 0.84 (0.31, 2.32) 0.74 1.16 (0.64, 2.12) 0.62 1.12 (0.29, 4.23) 0.38 1.96 (0.49, 7.86) 0.34 1.46 (0.56, 3.83) 0.44
IL10−819C>T 1.17 (0.58, 2.34) 0.65 0.80 (0.30, 2.16) 0.66 1.03 (0.59, 1.82) 0.91 0.77 (0.21, 2.74) 0.51 0.78 (0.21, 2.93) 0.72 0.78 (0.31, 1.94) 0.59
IL10−592C>A 1.16 (0.57, 2.31) 0.68 0.80 (0.30, 2.16) 0.66 1.02 (0.58, 1.81) 0.93 0.77 (0.21, 2.74) 0.51 0.78 (0.21, 2.93) 0.72 0.78 (0.31, 1.94) 0.59
TNFA−308G>A 2.41 (1.17, 4.94) 0.02 1.63 (0.61, 4.40) 0.33 2.11 (1.18, 3.77) 0.01 1.57 (0.47, 5.23) 0.46 0.41 (0.08, 2.01) 0.27 0.96 (0.37, 2.51) 0.94
IL6−174G>C 0.87 (0.32, 2.34) 0.79 1.98 (0.56, 6.95) 0.29 1.19 (0.55, 2.59) 0.66 0.37 (0.06, 2.22) 0.24 0.52 (0.14, 2.01) 0.34 0.46 (0.16, 1.35) 0.16
IL1RN+2018T>C 0.54 (0.18, 1.54) 0.25 1.18 (0.44, 3.15) 0.74 0.83 (0.40, 1.75) 0.63 1.28 (0.17, 9.08) 0.99 2.83 (0.75, 10.64) 0.12 2.37 (0.79, 7.11) 0.12
IFNG+874A>T 0.95 (0.30, 2.93) 0.92 1.51 (0.43, 5.30) 0.52 1.17 (0.50, 2.70) 0.72 0.39 (0.06, 2.33) 0.31 0.50 (0.14, 1.83) 0.30 0.46 (0.16, 1.31) 0.15
CCL17−431C>T 0.74 (0.09, 5.58) 0.77 0.65 (0.09, 4.94) 0.68 0.69 (0.17, 2.90) 0.61 b b 6.97 (1.54, 31.46) 0.01 b b

a Dominant model.

b Patients with CCL17−431T allele had no events in the GELA cohort.

CHL, Hodgkin lymphoma; HR, hazard ratio; CI, confidence interval.

gr1

Fig. 1 Progression free survival of EBV negative classical Hodgkin lymphoma patients by TNFA−308G>A polymorphism in GELA and SPORE cohorts.

4. Discussion

In these two prospective cohorts, we report that TNFA−308G>A was associated with PFS in EBV CHL, and these results were independent of the IPS. To our knowledge, no previous study has extensively assessed the relation between immune SNPs and CHL prognosis by integrating EBV status of the tumor. The prognostic value of TNFA−308G>A SNP was previously observed in diffuse large B-cell lymphoma (DLBCL) patients showing that TNFA−308A is associated with poor patient outcome [23] and [24] in accordance with our findings observed in EBV CHL.

EBV+ and EBV CHL appear to represent distinct entities with respect to differences among clinical characteristics, outcome [36] biologic features [41] and etiology [42] . Genetic risk factors differ among EBV+ and EBV CHL, genetic variants within the HLA class I region were found to be more frequently associated with EBV+ CHL and those in class II region with EBV CHL [34] . These results are in accordance with findings showing that T-cell cytotoxic response against EBV is predominantly driven by HLA class I molecules. These data highlight the potential importance of genetic background in the development of either EBV+ or EBV CHL, and raises the hypothesis that different host immune SNPs are involved in the control of the EBV+ and EBV tumors and consequently affect the prognosis of patients differently. In addition, biologic studies have shown that immunologic response was different in EBV+ and EBV CHL with a Th1/antiviral response in EBV+ CHL [35] with some specificities for cytokine or chemokine profiles observed in EBV+ CHL [43] . In this context, the type of immune SNPs that modulate cytokine expression should be different in EBV+ and EBV CHL.

While we observed no significant results for EBV+ CHL, a major limitation is the small number of patients and events in this group (i.e., a total of 76 cases for the two cohorts combined). Our EBV results were derived from exploratory analyses, and we did not adjust of multiple testing. Thus, further studies are needed to replicate our results in EBV CHL as well as evaluate these associations in a sufficiently powered study of EBV+ CHL.

Analyses for the combined GELA and SPORE cohorts found no correlation between PFS and any of the nine candidate SNPs, confirming the results of previous studies for IL10[28] and [32]IL6 [30] and TNFA [31] . One study of 184 HL patients (46 in tumor tissue samples) observed an effect of IL10−592C>A and IL6−174G>C for freedom from treatment failure (FFTF), with IL10−592AA and IL6−174GG at higher risk of relapse [31] . A recent study on 301 HL patients confirmed the result for the IL10−592 SNP, showing that patients with IL10−592AA, IL10−819TT or IL10−1082AA genotypes had an inferior FFTF [33] . In DLBCL, the prognostic impact of promoter IL10 SNPs is also conflicting [24], [25], [26], [27], and [28]. To our knowledge, the prognostic role of SNPs in the IL1RN, INFG and CCL17 genes for CHL patients have not been investigated previously, and we did not observe an association between these SNPs and patient outcome in the two cohorts.

The TNFA variant rs1800629 showed a difference of MAF between the GELA (0.125) and the SPORE (0.199) cohorts, and this remained similar for EBV and EBV+ CHL in each cohort. Technical error in genotyping is not a likely explanation, given the extensive quality controls. Cytokine SNPs can vary by race/ethnicity. In one population-based study, the TNFA MAF between African–American and Caucasian healthy women were 0.14 and 0.17, respectively [44] . In a NHL susceptibility study, TNFA MAFs were 0.18, 0.15, 0.10, 0.06 in White, Black, Hispanic and Asian NHL patients, respectively [22] . Unfortunately, we could not obtain data on the ethnicity by self-report in the French cohort (not allowed by law), so the role of racial/ethnic mixture on the French cohort is not known, but is not likely to be substantial since most patients are likely to be of Western-European ancestry. In one previous clinical study conducted in France for DLBCL patients, MAF of this variant was 0.17 as compared to 0.16 for healthy French controls [45] and 0.19 for DLBCL patients included in a USA study [24] . Thus, the differences in MAF between the two studies, which are quite modest and are in the range of prior reports, may just be the result of sampling variability.

In this analysis, we also assessed the association of SNPs from cytokine genes described as functional in previous in vitro or in vivo studies [17] with circulating cytokines in the context of CHL. Globally, we did not observe a correlation between each SNP genotype and the median level of the related cytokines, either in GELA (plasma) or SPORE (serum) cohorts. The correlation between genotypes and cytokine levels has been mostly derived from in vitro studies, and cytokines were generally measured in the peripheral blood mononuclear cell (PBMC) supernatant with different leukocyte stimulation assays [17] . We could not confirm prior studies in CHL showing that IL6−174G>C, IL10−592C>A and IL10−1082A>G were associated with related cytokine levels [46] and [47], or that a higher TNFα cytokine production is observed in NHL patients with TNFA−308A allele [23] . In fact, regarding our significant outcome results for TNFA−308G>A in EBV CHL, TNFA−308A and TNFA−308GG patients had similar TNFα levels in two different analyzed supernatants.

One hypothesis to explain the difference of prognosis between TNFA−308A and TNFA−308GG patients is that the variant may influence the cytokine variability in the tumor microenvironment that is not reflected by the circulating TNFα levels. At a molecular level, SNPs in TNFA promoter could influence the binding of transcriptional factors as NF-κB p50–p50 dimers or OCT-1 with potential consequences on gene expression [48] and [49]. In the same chromosome 6 locus, there is a functional SNP in the promoter of inhibitory kappaB-like gene, IκBL−63A>T (rs2071592) in complete linkage of disequilibrium with TNFA−308G>A [50] . However, there are no data whether the IκBL protein has a similar function to IκB. The HRS cell is characterized by a global cellular reprogramming with the deregulation of many transcriptional factors [51] and EBV participates to the down-regulation of the B-cell phenotype in HRS [52] and [53]. Interestingly, a recent genome-wide transcriptional analysis of micro-dissected HRS cells demonstrated that relatively few transcriptional changes were driven by the presence of EBV in HRS cells, leading the authors to hypothesize that EBV might influence gene expression in the tumor microenvironment rather than impacting the HRS cells [54] as observed by Chetaille et al. with a Th1/antiviral response in EBV+ CHL [35] . In this context, it will be interesting to study whether germline SNPs in immune mediators affect the composition of reactive cells of tumor microenvironment, which is known to influence prognosis of patients [16] and [55] and if EBV status further impacts these findings.

5. Conclusions

We found in two prospective cohorts of representative CHL patients that SNPs in the promoter of TNFA influenced progression free survival in patients with EBV CHL, independent of IPS. Our study was limited by the small number of patients evaluated for EBV+ CHL, and lack of a controlled treatment trial setting for the SPORE. However, the two cohorts were very similar from a clinical perspective, and should be highly generalizable to CHL in Caucasian patients. The cytokine targets chosen in this study, although previously identified in several reports as good candidates for prognostication of CHL patients, are unlikely to fully reflect the complex network of immune mediators involved in the pathogenesis of CHL. Given the complexity of immune interactions in CHL, further studies using whole-genome scans [34] and [56] or sequencing in large cohorts are necessary to determine the prognostic impact of inheritable factors in this disease, integrating data related to EBV status with subsequently functional studies.

Acknowledgments

The authors thank the GELARC (Groupe d’Etude des Lymphomes de l’Adultes Recherche Clinique) for the clinical management of this work.

This work was supported by a grant from the French Ministry of Health (PHRC 1998 and 2002); la Fondation de France (postdoctoral fellowship, H.G.); Philippe Foundation (postdoctoral fellowship, H.G.); Grants CA92104, CA92153, CA97274, and CA25224 from the US National Institutes of Health; and by the Predolin Foundation.

References

  • [1] D. Hasenclever, V. Diehl. A prognostic score for advanced Hodgkin’s disease. International Prognostic Factors Project on Advanced Hodgkin’s Disease. N Engl J Med. 1998;339:1506-1514 Crossref.
  • [2] C. Steidl, A. Diepstra, T. Lee, F.C. Chan, P. Farinha, K. Tan, et al. Gene expression profiling of microdissected Hodgkin Reed-Sternberg cells correlates with treatment outcome in classical Hodgkin lymphoma. Blood. 2012;120:3530-3540 Crossref.
  • [3] B.F. Skinnider, T.W. Mak. The role of cytokines in classical Hodgkin lymphoma. Blood. 2002;99:4283-4297 Crossref.
  • [4] E.D. Hsi. Biologic features of Hodgkin lymphoma and the development of biologic prognostic factors in Hodgkin lymphoma: tumor and microenvironment. Leuk Lymphoma. 2008;49:1668-1680 Crossref.
  • [5] R.O. Casasnovas, N. Mounier, P. Brice, M. Divine, F. Morschhauser, J. Gabarre, et al. Plasma cytokine and soluble receptor signature predicts outcome of patients with classical Hodgkin’s lymphoma: a study from the Groupe d’Etude des Lymphomes de l’Adulte. J Clin Oncol. 2007;25:1732-1740 Crossref.
  • [6] U. Axdorph, J. Sjöberg, G. Grimfors, O. Landgren, A. Porwit-MacDonald, M. Björkholm. Biological markers may add to prediction of outcome achieved by the International Prognostic Score in Hodgkin’s disease. Ann Oncol. 2000;11:1405-1411 Crossref.
  • [7] G. Nadali, L. Tavecchia, E. Zanolin, V. Bonfante, S. Viviani, E. Camerini, et al. Serum level of the soluble form of the CD30 molecule identifies patients with Hodgkin’s disease at high risk of unfavorable outcome. Blood. 1998;91:3011-3016
  • [8] A.H. Sarris, K.O. Kliche, P. Pethambaram, A. Preti, S. Tucker, C. Jackow, et al. Interleukin-10 levels are often elevated in serum of adults with Hodgkin’s disease and are associated with inferior failure-free survival. Ann Oncol. 1999;10:433-440 Crossref.
  • [9] T.P. Vassilakopoulos, G. Nadali, M.K. Angelopoulou, M.P. Siakantaris, M.N. Dimopoulou, F.N. Kontopidou, et al. Serum interleukin-10 levels are an independent prognostic factor for patients with Hodgkin’s lymphoma. Haematologica. 2001;86:274-281
  • [10] J.F. Seymour, M. Talpaz, F.B. Hagemeister, F. Cabanillas, R. Kurzrock. Clinical correlates of elevated serum levels of interleukin 6 in patients with untreated Hodgkin’s disease. Am J Med. 1997;102:21-28
  • [11] S.M. Ansell, M.J. Maurer, S.C. Ziesmer, S.L. Slager, T.M. Habermann, B.K. Link, et al. Pretreatment serum cytokines predict early disease relapse and a poor prognosis in newly diagnosed classical Hodgkin lymphoma (cHL) patients. ASH Ann Meet Abstr. 2011;118:429
  • [12] K. Warzocha, J. Bienvenu, P. Ribeiro, I. Moullet, C. Dumontet, E.M. Neidhardt-Berard, et al. Plasma levels of tumour necrosis factor and its soluble receptors correlate with clinical features and outcome of Hodgkin’s disease patients. Br J Cancer. 1998;77:2357-2362
  • [13] C. Tecchio, G. Nadali, P. Scapini, C. Bonetto, C. Visco, N. Tamassia, et al. High serum levels of B-lymphocyte stimulator are associated with clinical–pathological features and outcome in classical Hodgkin lymphoma. Br J Haematol. 2007;137:553-559 Crossref.
  • [14] M.R. Weihrauch, O. Manzke, M. Beyer, H. Haverkamp, V. Diehl, H. Bohlen, et al. Elevated serum levels of CC thymus and activation-related chemokine (TARC) in primary Hodgkin’s disease: potential for a prognostic factor. Cancer Res. 2005;65:5516-5519 Crossref.
  • [15] W.J. Plattel, A. van den Berg, L. Visser, A.M. van der Graaf, J. Pruim, H. Vos, et al. Plasma thymus and activation-regulated chemokine as an early response marker in classical Hodgkin’s lymphoma. Haematologica. 2012;97:410-415 Crossref.
  • [16] C. Steidl, T. Lee, S.P. Shah, P. Farinha, G. Han, T. Nayar, et al. Tumor-associated macrophages and survival in classic Hodgkin’s lymphoma. N Engl J Med. 2010;362:875-885 Crossref.
  • [17] A.J. Smith, S.E. Humphries. Cytokine and cytokine receptor gene polymorphisms and their functionality. Cytokine Growth Factor Rev. 2009;20:43-59 Crossref.
  • [18] N. Rothman, C.F. Skibola, S.S. Wang, G. Morgan, Q. Lan, M.T. Smith, et al. Genetic variation in TNF and IL10 and risk of non-Hodgkin lymphoma: a report from the InterLymph Consortium. Lancet Oncol. 2006;7:27-38 Crossref.
  • [19] Q. Lan, T. Zheng, N. Rothman, Y. Zhang, S.S. Wang, M. Shen, et al. Cytokine polymorphisms in the Th1/Th2 pathway and susceptibility to non-Hodgkin lymphoma. Blood. 2006;107:4101-4108 Crossref.
  • [20] M.P. Purdue, Q. Lan, A. Kricker, A.E. Grulich, C.M. Vajdic, J. Turner, et al. Polymorphisms in immune function genes and risk of non-Hodgkin lymphoma: findings from the New South Wales non-Hodgkin Lymphoma Study. Carcinogenesis. 2007;28:704-712
  • [21] S.S. Wang, M.P. Purdue, J.R. Cerhan, T. Zheng, I. Menashe, B.K. Armstrong, et al. Common gene variants in the tumor necrosis factor (TNF) and TNF receptor superfamilies and NF-kB transcription factors and non-Hodgkin lymphoma risk. PLoS One. 2009;4:e5360 Crossref.
  • [22] C.F. Skibola, P.M. Bracci, A. Nieters, A. Brooks-Wilson, S. de Sanjosé, A.M. Hughes, et al. Tumor necrosis factor (TNF) and lymphotoxin-alpha (LTA) polymorphisms and risk of non-Hodgkin lymphoma in the InterLymph Consortium. Am J Epidemiol. 2010;171:267-276 Crossref.
  • [23] K. Warzocha, P. Ribeiro, J. Bienvenu, P. Roy, C. Charlot, D. Rigal, et al. Genetic polymorphisms in the tumor necrosis factor locus influence non-Hodgkin’s lymphoma outcome. Blood. 1998;91:3574-3581
  • [24] T.M. Habermann, S.S. Wang, M.J. Maurer, L.M. Morton, C.F. Lynch, S.M. Ansell, et al. Host immune gene polymorphisms in combination with clinical and demographic factors predict late survival in diffuse large B-cell lymphoma patients in the pre-rituximab era. Blood. 2008;112:2694-2702 Crossref.
  • [25] E. Lech-Maranda, L. Baseggio, J. Bienvenu, C. Charlot, F. Berger, D. Rigal, et al. Interleukin-10 gene promoter polymorphisms influence the clinical outcome of diffuse large B-cell lymphoma. Blood. 2004;103:3529-3534 Crossref.
  • [26] M. Berglund, U. Thunberg, G. Roos, R. Rosenquist, G. Enblad. The interleukin-10 gene promoter polymorphism (−1082) does not correlate with clinical outcome in diffuse large B-cell lymphoma. Blood. 2005;105:4894-4895 Crossref.
  • [27] D. Kube, T.D. Hua, M. Kloss, B. Kulle, J. Brockmoller, L. Wojnowski, et al. The interleukin-10 gene promoter polymorphism −1087AG does not correlate with clinical outcome in non-Hodgkin’s lymphoma. Genes Immun. 2007;8:164-167 Crossref.
  • [28] E. Domingo-Domenech, Y. Benavente, E. Gonzalez-Barca, C. Montalban, J. Guma, R. Bosch, et al. Impact of interleukin-10 polymorphisms (−1082 and −3575) on the survival of patients with lymphoid neoplasms. Haematologica. 2007;92:1475-1481 Crossref.
  • [29] J.R. Cerhan, S. Wang, M.J. Maurer, S.M. Ansell, S.M. Geyer, W. Cozen, et al. Prognostic significance of host immune gene polymorphisms in follicular lymphoma survival. Blood. 2007;109:5439-5446 Crossref.
  • [30] P. Cordano, A. Lake, L. Shield, G.M. Taylor, F.E. Alexander, P.R. Taylor, et al. Effect of IL-6 promoter polymorphism on incidence and outcome in Hodgkin’s lymphoma. Br J Haematol. 2005;128:493-495 Crossref.
  • [31] S. Hohaus, M. Giachelia, A. Di Febo, M. Martini, G. Massini, B. Vannata, et al. Polymorphism in cytokine genes as prognostic markers in Hodgkin’s lymphoma. Ann Oncol. 2007;18:1376-1381 Crossref.
  • [32] O.E. Yri, P.O. Ekstrom, V. Hilden, G. Gaudernack, K. Liestol, E.B. Smeland, et al. Polymorphisms in genes encoding interleukin-10 and drug metabolizing enzymes GSTP1, GSTT1, GSTA1 and UGT1A1 influence risk and outcome in Hodgkin lymphoma. Leuk Lymphoma. 2012;53:1934-1944 Crossref.
  • [33] N. Schoof, J. Franklin, R. Furst, T. Zander, B.F. Von, F. Peyrade, et al. Interleukin-10 gene polymorphisms are associated with freedom from treatment failure for patients with Hodgkin lymphoma. Oncologist. 2013;18:80-89 Crossref.
  • [34] K.Y. Urayama, R.F. Jarrett, H. Hjalgrim, A. Diepstra, Y. Kamatani, A. Chabrier, et al. Genome-wide association study of classical Hodgkin lymphoma and Epstein-Barr virus status-defined subgroups. J Natl Cancer Inst. 2012;104:240-253 Crossref.
  • [35] B. Chetaille, F. Bertucci, P. Finetti, B. Esterni, A. Stamatoullas, J.M. Picquenot, et al. Molecular profiling of classical Hodgkin lymphoma tissues uncovers variations in the tumor microenvironment and correlations with EBV infection and outcome. Blood. 2009;113:2765-3775 Crossref.
  • [36] A. Diepstra, G.W. van Imhoff, M. Schaapveld, H. Karim-Kos, B.A. Van den, E. Vellenga, et al. Latent Epstein-Barr virus infection of tumor cells in classical Hodgkin’s lymphoma predicts adverse outcome in older adult patients. J Clin Oncol. 2009;27:3815-3821 Crossref.
  • [37] M. Sanson, F.J. Hosking, S. Shete, D. Zelenika, S.E. Dobbins, Y. Ma, et al. Chromosome 7p11.2 (EGFR) variation influences glioma risk. Hum Mol Genet. 2011;20:2897-2904 Crossref.
  • [38] C.A. Thompson, M.J. Maurer, J.R. Cerhan, J.A. Katzmann, S.M. Ansell, T.M. Habermann, et al. Elevated serum free light chains are associated with inferior event free and overall survival in Hodgkin lymphoma. Am J Hematol. 2011;86:998-1000 Crossref.
  • [39] Y. Tsunemi, M. Komine, T. Sekiya, H. Saeki, K. Nakamura, K. Hirai, et al. The −431C>T polymorphism of thymus and activation-regulated chemokine increases the promoter activity but is not associated with susceptibility to atopic dermatitis in Japanese patients. Exp Dermatol. 2004;13:715-719 Crossref.
  • [40] M.K.B. Parmar, V. Torri, L. Stewart. Extracting summary statistics to perform meta-analyses of the published literature for survival endpoints. Stat Med. 1998;17:2815-2834 Crossref.
  • [41] G. Kapatai, P. Murray. Contribution of the Epstein Barr virus to the molecular pathogenesis of Hodgkin lymphoma. J Clin Pathol. 2007;60:1342-1349
  • [42] H. Hjalgrim. On the aetiology of Hodgkin lymphoma. Dan Med J. 2012;59:B4485
  • [43] D. Aldinucci, A. Gloghini, A. Pinto, R. De Filippi, A. Carbone. The classical Hodgkin’s lymphoma microenvironment and its role in promoting tumour growth and immune escape. J Pathol. 2010;221:248-263 Crossref.
  • [44] A.L. Van Dyke, M.L. Cote, A.S. Wenzlaff, S. Land, A.G. Schwartz. Cytokine SNPs: comparison of allele frequencies by race and implications for future studies. Cytokine. 2009;46:236-244 Crossref.
  • [45] P. Juszczynski, E. Kalinka, J. Bienvenu, G. Woszczek, M. Borowiec, T. Robak, et al. Human leukocyte antigens class II and tumor necrosis factor genetic polymorphisms are independent predictors of non-Hodgkin lymphoma outcome. Blood. 2002;100:3037-3040 Crossref.
  • [46] W. Cozen, P.S. Gill, S.A. Ingles, R. Masood, O. Martinez-Maza, M.G. Cockburn, et al. IL-6 levels and genotype are associated with risk of young adult Hodgkin lymphoma. Blood. 2004;103:3216-3221 Crossref.
  • [47] S. Hohaus, M. Giachelia, G. Massini, B. Vannata, M. Criscuolo, M. Martini, et al. Clinical significance of interleukin-10 gene polymorphisms and plasma levels in Hodgkin lymphoma. Leuk Res. 2009;33:1352-1356 Crossref.
  • [48] I.A. Udalova, A. Richardson, A. Denys, C. Smith, H. Ackerman, B. Foxwell, et al. Functional consequences of a polymorphism affecting NF-kappaB p50–p50 binding to the TNF promoter region. Mol Cell Biol. 2000;20:9113-9119 Crossref.
  • [49] J.C. Knight, I. Udalova, A.V. Hill, B.M. Greenwood, N. Peshu, K. Marsh, et al. A polymorphism that affects OCT-1 binding to the TNF promoter region is associated with severe malaria. Nat Genet. 1999;22:145-150 Crossref.
  • [50] R.J. Allcock, K. Baluchova, K.Y. Cheong, P. Price. Haplotypic single nucleotide polymorphisms in the central MHC gene IKBL, a potential regulator of NF-kappaB function. Immunogenetics. 2001;52:289-293
  • [51] R. Kuppers. The biology of Hodgkin’s lymphoma. Nat Rev Cancer. 2009;9:15-27 Crossref.
  • [52] M. Vockerodt, S.L. Morgan, M. Kuo, W. Wei, M.B. Chukwuma, J.R. Arrand, et al. The Epstein-Barr virus oncoprotein, latent membrane protein-1, reprograms germinal centre B cells towards a Hodgkin’s Reed-Sternberg-like phenotype. J Pathol. 2008;216:83-92 Crossref.
  • [53] T. Portis, P. Dyck, R. Longnecker. Epstein-Barr Virus (EBV) LMP2A induces alterations in gene transcription similar to those observed in Reed-Sternberg cells of Hodgkin lymphoma. Blood. 2003;102:4166-4178 Crossref.
  • [54] E. Tiacci, C. Doring, V. Brune, C.J. van Noesel, W. Klapper, G. Mechtersheimer, et al. Analyzing primary Hodgkin and Reed-Sternberg cells to capture the molecular and cellular pathogenesis of classical Hodgkin lymphoma. Blood. 2012;120:4609-4620 Crossref.
  • [55] P. Greaves, A. Clear, R. Coutinho, A. Wilson, J. Matthews, A. Owen, et al. Expression of FOXP3, CD68, and CD20 at diagnosis in the microenvironment of classical Hodgkin lymphoma is predictive of outcome. J Clin Oncol. 2012;10:256-262
  • [56] V. Enciso-Mora, P. Broderick, Y. Ma, R.F. Jarrett, H. Hjalgrim, K. Hemminki, et al. A genome-wide association study of Hodgkin’s lymphoma identifies new susceptibility loci at 2p16.1 (REL), 8q24.21 and 10p14 (GATA3). Nat Genet. 2010;42:1126-1130 Crossref.

Footnotes

a Onco-Hematology, Centre Léon Bérard, UMR CNRS 5239, Université Lyon 1, Lyon, France

b Health Sciences Research, Mayo Clinic, Rochester, MN, USA

c Department of Hematology, CHU Dijon, Dijon, France

d Division of Hematology, Department of Internal Medicine, Mayo Clinic, Rochester, MN, USA

e Department of Hematology, Warsaw, Poland

f UMR CNRS 5239, Université Lyon 1, Lyon, France

g Department of Hematology, APHP, Hôpital Saint Louis, Paris, France

h Department of Hematology, CHU de Besançon, France

i Department of Hematology, CHU, Lille, France

j Unité Hémopathies lymphoïdes, Hôpital Henri Mondor, Créteil, France

k University of Iowa, Holden Comprehensive Cancer Center, Iowa City, IA, USA

l Department of Hematology, Centre Henri Becquerel, Rouen, France

m Department of Hematology, Institut Gustave Roussy, Villejuif, France

n Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, USA

o Department of Pathology, APHP, Hôtel Dieu, Paris, France

p Department of Hematology, UMR CNRS 5239, Université Lyon 1, HCL, Pierre-Bénite, France

lowast Corresponding author. Address: Centre Léon Bérard, UMR CNRS 5239, Université Lyon 1, 28, rue Laennec, 69008 Lyon, France. Tel.: +33 4 78 78 26 41; fax: +33 4 78 78 27 16.