Association of IL6 and CCL2 gene polymorphisms with the ... - Nature

Bone Marrow Transplantation (2009) 44, 227–235 & 2009 Macmillan Publishers Limited All rights reserved 0268-3369/09 $32.00

www.nature.com/bmt

ORIGINAL ARTICLE

Association of IL6 and CCL2 gene polymorphisms with the outcome of allogeneic haematopoietic stem cell transplantation Z Ambruzova1,6, F Mrazek1,6, L Raida2, P Jindra3, B Vidan-Jeras4, E Faber2, J Pretnar5, K Indrak2 and M Petrek1 1 Laboratory of Immunogenomics, Department of Immunology, Palacky University and University Hospital, Olomouc, Czech Republic; 2Department of Haemato-Oncology, Palacky University and University Hospital, Olomouc, Czech Republic; 3Department of Haematology and Oncology, University Hospital, Pilsen, Czech Republic; 4Tissue Typing Centre, Blood Transfusion Centre of Slovenia, Ljubljana, Slovenia and 5Department of Haematology, University Medical Centre, Ljubljana, Slovenia

Various polymorphisms of non-HLA genes have recently been investigated as candidate risk factors in allogeneic haematopoietic SCT (aHSCT). Our study aimed at exploring possible associations of IL6 and CCL2 single nucleotide polymorphisms (SNP) with aHSCT outcome. A total of 166 HLA-identical aHSCT pairs recruited in were genotyped for IL6 174 G/C, IL6 597 G/A, CCL2 2518 A/G and CCL2 2076 A/T SNPs by PCR with sequence-specific primers (PCR-SSP). The association between IL6 174 GG genotype and increased risk of acute GVHD was found in whole study group (P ¼ 0.03) and in the subgroup of related aHSCT (P ¼ 0.01), association between IL6 597 GG genotype and the occurrence of acute GVHD was detected only in the related aHSCT pairs (P ¼ 0.02). Furthermore, reduction in OS was revealed among recipients possessing IL6 174*G allele in the group of related aHSCT pairs (P ¼ 0.04). Presence of CCL2 2076 TT genotype was associated with decrease of OS (P ¼ 0.04) and increase of TRM (P ¼ 0.02) in patients transplanted by related donor. These results, in the context of previous findings, suggest that IL6 gene polymorphisms may be associated with aHSCT outcome, particularly in patients transplanted from a related donor. Bone Marrow Transplantation (2009) 44, 227–235; doi:10.1038/bmt.2009.16; published online 23 February 2009 Keywords: IL-6; CCL2; single nucleotide polymorphism; haematopoietic SCT; GVHD; OS

Correspondence: Dr M Petrek, Laboratory of Immunogenomics, Department of Immunology, Palacky University and University Hospital Olomouc, IP Pavlova 6, Olomouc CZ-775 20, Czech Republic. E-mail: [email protected] 6 These authors contributed equally to this work. Received 30 April 2008; revised 8 January 2009; accepted 11 January 2009; published online 23 February 2009

Introduction Allogeneic haematopoietic SCT (aHSCT) has evolved over the last 50 years to become an important therapeutic approach, routinely used in the treatment of haematological malignancies and also in many other non-malignant diseases.1 However, although the overall outcome of aHSCT has dramatically improved in the past decade, it is still compromised by the development of serious posttransplant complications such as GVHD or infections resulting in significant TRM. Even with precise HLA matching between donor and recipient, other genetic factors (non-HLA polymorphisms) are thought to influence the outcome of transplantation.2,3 In this regard, single nucleotide polymorphisms (SNP) of the genes for cytokines/chemokines and their receptors, minor histocompatibility Ag, genes for innate immunity and those associated with the metabolism of drugs have recently been studied for their possible association with aHSCT outcome, often with controversial results.4–6 Of non-HLA factors, cytokines are important candidates in searching for marker of GVHD. IL-6 is an important proinflammatory cytokine implicated in immune-mediated complications of aHSCT, especially GVHD. IL-6 production may be affected by functional SNP in the promoter region of the IL6 gene at position 174 (G/C). Healthy donors possessing the G allele produce higher serum levels of IL-6.7 IL6 174*G allele in patient and/or donor has been found to be associated with the risk of development of acute or chronic GVHD in previous studies.5,8,9 Monocyte chemoattractant protein-1 (MCP-1, systematic name CCL2) belongs to the subfamily of CC chemotactic cytokines (chemokines).10 It attracts the mononuclear cells to the site of inflammation. In the aHSCT context, MCP-1 may potentiate migration of donor T cells during the activation phase of GVHD.11 Several polymorphisms have been identified in the promoter region of CCL2 gene, of which the polymorphisms 2518 G/A and 2076 A/T seem to have a functional effect at least in vitro.12 In the context of organ transplantation, determination of CCL2 overproduction was showed associated with significant deterioration of kidney graft

IL6 and CCL2 gene polymorphisms in aHSCT Z Ambruzova et al

228

function.13 In contrast, no association of CCL2 2518 gene variants and acute rejection in long-term survival after liver transplantation was found.14 In this study, IL6 gene polymorphisms at positions 174 (G/C) and 597 (G/A) and CCL2 gene polymorphisms at positions 2518 (A/G) and 2076 (A/T) were analysed in subcohorts of related and unrelated HLA-identical aHSCT donor–recipient pairs. These aforementioned pairs of IL6 and CCL2 gene polymorphisms were selected because of their localization in the regulatory regions (promoters) of IL6 and CCL2 genes together with their documented or postulated effect on expression of encoded mediators. The potential genetic associations of these selected SNPs with the development of acute or chronic GVHD, TRM and OS after aHSCT were screened in the whole cohort of the aHSCT pairs and subsequently in the context of related and unrelated aHSCT.

Materials and methods Study population and design of the study A total of 166 HLA-identical aHSCT pairs from three different centres were included in this study: 88 aHSCT pairs were recruited in Olomouc, Czech Republic (centre 1), further 78 aHSCT pairs were then added into the study from the centres of Pilsen, Czech Republic (centre 2; n ¼ 25) and Ljubljana, Slovenia (centre 3; n ¼ 53). All patients received graft from HLA-identical donor. In case of related aHSCT pairs, all donors were genotypically identical; the adequate level of testing according the EFI standards was used for definitive establishing of HLA identity (HLA-A, -B, -DR loci as minimum). In case of unrelated aHSCT pairs, only those with complete identity at allelic (‘high resolution’) level of HLA-A,-B,-Cw,-DRB1 and -DQB1 loci were involved. All patients signed in their centres the informed consent on the usage of DNA samples taken primarily for diagnostic purposes (HLA genotyping) and also their anonymous clinical and laboratory data for the research purposes. The study was approved by the Ethics Committee of the Medical Faculty Palacky University and Faculty Hospital Olomouc and the local ethical authorities in Pilsen (University Hospital) and Ljubljana (University Medical Centre). The whole study group was divided into two subcohorts. The first subcohort consisted of 121 patients receiving graft from related donor and the second subcohort included 45 patients transplanted using voluntary unrelated donors. Treatment, conditioning regimen and GVHD patient prophylaxis was performed according to the local transplantation protocols. Detailed characteristics of aHSCT patient/donor pairs are given in Table 1. To enable us to compare the frequencies of selected SNPs in the groups of aHSCT patients with those observed in general population, the IL6 and CCL2 genotyping data representative for the Czech healthy population have been obtained on the population sample of the unrelated healthy subjects of the Czech origin used for the purposes of the article by Kubistova et al.15 after enlargement of the population group up to 135 subjects. Bone Marrow Transplantation

Genetic analysis DNA was extracted from the peripheral blood of the patients and their donors by the standard salting out method or with the use of Qiagen kit (QiAmp Blood kit; Qiagen, Hilden, Germany) and stored at 20 1C. Sequence specific primers used for determination of IL6 174 (G/C, rs1800795), IL6 597 (G/A, rs1800797), CCL2 2076 (A/T, rs1024610) and CCL2 2518 (A/G, rs1024611) genotypes by PCR are listed in online Appendix 1. Reaction conditions and internal control primers were adopted from the Phototyping methodology.16 The development of in house PCR-SSP genotyping protocol for both investigated IL6 SNPs has been controlled by parallel typing of randomly selected control samples of all genotypes by another validated genotyping assay (Cytokine Typing Tray kit, University of Heidelberg, Heidelberg, Germany). In case of CCL2 SNPs, the protocol has been validated by the ‘complementary’ designed PCR-SSP assay (that is, by alternative protocol with opposite direction of specific primers). Complete concordance of genotype calling obtained by main (executive) and checking protocols was observed for all investigated SNPs.

Statistics The genotypes for IL6 and CCL2 SNPs were determined by direct counting and appropriate allele and phenotype (carriage rates) frequencies were calculated. Distribution of genotypes was tested for conformity with the Hardy– Weinberg equilibrium using the w2-test. IL6 and CCL2 haplotype analysis was performed using software Arlequin 3.0 (University of Berne, Switzerland, http://cmpg.unibe. ch/software/arlequin3).17 Associations between patient or donor IL6 and CCL2 genotypes/phenotypes for all SNPs and main outcome of the study, development of acute or chronic GVHD for subcohort I, subcohort II and both cohorts together were analysed using the software SPSS 15.0 (SPSS Inc., Chicago, IL, USA). For univariate analysis Pearson’s w2-test was used, the multivariate analysis was performed using binary logistic regression models; factors biologically considerable in the occurrence of GVHD were included (diagnosis, stem cell source, conditioning regimen, GVHD prophylaxis, donor–recipient sex combination (female donor in male recipients vs others) and IL6–CCL2 variants). Patients who died before day þ 30 after transplantation without significant aGVHD symptoms were excluded from acute GVHD association analysis and similarly patients who died before day þ 100 after transplantation not evaluable for cGVHD were excluded from chronic GVHD association analysis. The Kaplan–Meier and Cox regression analysis for OS and TRM was performed using SPSS software (version 15.0). In addition to factors mentioned above the Cox regression analysis of OS included also acute and chronic GVHD data. TRM was defined as death from any cause, other than progression of the underlying disease during the course of treatment. P values less than 0.05 were considered statistically significant.


229 Table 1

Characteristics of the aHSCT patient/donor pairs according to cohorts

Characteristic

All aHSCT

Subcohort I—related aHSCT

Subcohort II—unrelated aHSCT

Number of patients Age (median, range) (years) Sex (female/male)

166 43 (17–64) 72/94

121 44 (17–64) 48/73

45 40 (19–61) 24/21

Diagnosis AML ALL CML NHL CLL Other

75 21 19 14 7 30

59 10 15 11 6 20

16 11 4 3 1 10

Disease risk Low High Not determined

62 51 53

41 40 40

21 11 13

Gender compatibility Matched Female donor/male recipient Male donor/female recipient

86 37 43

62 31 28

24 6 15

Conditioning regimen Myeloablative BU+CY±ATG CY+TBI Other combination Non-myeloablative Fludarabine+CY±ATG Fludarabine+melfalan Other combination

96 55 20 21 70 18 16 36

67 37 16 14 54 13 11 30

29 18 4 7 16 5 5 6

GVHD prophylaxis CsA only CsA+MTX CsA+MMF Other combination No prophylaxis

27 97 22 18 2

23 74 14 8 2

4 23 8 10 0

Type of donor Related Unrelated

121 45

121 0

0 45

Source of stem cells BM PBSC

22 144

16 105

6 39

Acute GVHD Grade 0–I Grade II Grade III Grade IV

107 40 10 9

79 31 6 5

28 9 4 4

Chronic GVHD 0 Limited Extensive Not determined

108 32 25 1

78 21 21 1

30 11 4 0

78 88 25 63

54 67 23 44

24 21 2 19

Survival status Alive Dead GVHD related cause Non-GVHD cause

Abbreviations: aHSCT ¼ allogeneic haematopoietic SCT; ATG ¼ antithymocyte globulin; MMF ¼ mycophenolate mofetil; NHL ¼ non-Hodgkin’s lymphoma.

Bone Marrow Transplantation


230 80%

Table 2 Genotype frequencies of the IL6 174 G/C, IL6 597 G/A, CCL2 2518 A/G and CCL2 2076 A/T SNPs in patients, donors and healthy control group Patients n ¼ 166

Donors n ¼ 166

Controls n ¼ 135

IL6 -174 Genotype GG Genotype GC Genotype CC

0.28 (47) 0.49 (82) 0.23 (37)

0.22 (37) 0.56 (93) 0.22 (35)

0.33 (45) 0.48 (65) 0.19 (25)

IL6 -597 Genotype GG Genotype GA Genotype AA

0.30 (49) 0.49 (80) 0.21 (34)

0.24 (39) 0.55 (89) 0.21 (35)

0.33 (45) 0.48 (65) 0.19 (25)

CCL2 –2518 Genotype AA Genotype AG Genotype GG

0.58 (96) 0.38 (63) 0.04 (6)

0.50 (82) 0.42 (70) 0.08 (13)

0.52 (68) 0.40 (52) 0.08 (11)

CCL2 –2076 Genotype AA Genotype AT Genotype TT

0.59 (96)a 0.36 (60) 0.05 (8)

0.69 (113) 0.28 (46) 0.03 (4)

0.73 (96) 0.26 (34) 0.01 (1)

Frequency of aGVHD

SNP

IL6 - 174 GG

70%

IL6 - 174 GC or CC

17/30

60% 22/42 50%

5/12 40%

39/116

12/30

27/86

30% 20% 10% 0% All aHSCT

Related aHSCT

Unrelated aHSCT

Figure 1 Frequency of acute GVHD (grades II–IV) in the groups of patients with IL6 174 GG and recipients with other genotypes (IL6 174 GC or CC) analysed in whole study cohort, subcohorts I (related allogeneic haematopoietic SCT (aHSCT)) and II (unrelated aHSCT). Black bars represent recipients carrying IL6 174 GG genotype; grey bars patients with IL6 174 GC or CC genotypes. The frequency of aGVHD in IL6 174 GG homozygotes was increased in all analysed groups except of the subcohort II (P ¼ 0.03 in whole study cohort, P ¼ 0.01 in subcohort I—related aHSCT).

Abbreviation: SNP ¼ single nucleotide polymorphism. Data are presented as relative proportions of genotypes with the absolute numbers in parentheses. Distribution of genotypes in particular groups was in compliance with Hardy–Weinberg equilibrium (P40.05). a CCL2 2076 AA genotype frequency: P ¼ 0.04 for the comparison patients with donors, P ¼ 0.01 for the comparison patients with healthy controls.

1.0

Results General population data The distribution of the IL6 and CCL2 genotypes was in compliance with Hardy–Weinberg equilibrium in all groups of patients, donors and healthy controls (P40.05). Observed genotype frequencies for investigated SNPs were similar with those reported from other European centres/ populations (German, Czech, Polish, Bulgarian) referred by Kubistova15 and with the data listed in SNP database (http://www.ncbi.nlm.nih.gov/sites/entrez) for European populations. The frequencies of the IL6 174/597 haplotypes estimated by expectation–maximization algorithm (software Arlequin, version 3.0)17 are listed in online Appendix 2. As expected, strong (almost absolute) linkage disequilibrium was observed between the less common alleles of these two SNPs, that is, allele C of the 174 SNP and allele A of the 597 SNP in all study groups (patients and donors: D0 ¼ 0.99, healthy controls: D0 ¼ 0.98; Po0.00001 for all three groups). There were no significant differences between any of the groups for the genotype and allelic frequencies of IL6 174, IL6 597 and CCL2 2518 SNPs. The frequency of CCL2 2076 AA genotype was significantly lower in patients (59.0%) than in the group of their donors (69.0%; P ¼ 0.04) or healthy controls (73%; P ¼ 0.01; (Table 2). Analysis of IL6 gene polymorphisms association with aHSCT outcome IL6 gene variants in the whole aHSCT group. To investigate any possible association between IL6 gene Bone Marrow Transplantation

Overall survival

0.8

0.6 IL6 - 174 CC genotype 0.4

IL6 - 174 GG or GC genotype

0.2

0.0 0

20

40 60 80 Months after aHSCT

100

120

Figure 2 OS in relation to IL6 174 genotypes of the aHSCT recipient in the group of related aHSCT. A significant decrease of OS in patients possessing at least one IL6 174 G* allele (IL6 174 GG or GC genotype) compared to recipients with IL6 174 CC genotype was observed (log-rank test: P=0.04)

variants and aGVHD we compared the proportion of alleles and genotypes between the recipients with/without clinically significant aGVHD (graded II–IV). In the whole study group, IL6 174 GG homozygous patients developed aGVHD more frequently than individuals with the other genotypes: aGVHD was present in 22 out of 42 patients with GG genotype (52.4%) compared to 39 out of 116 patients with other genotypes (33.6%; P ¼ 0.03; OR 2.172; 95% confidence interval (CI) 1.060–4.451; Figure 1). Similarly, patients carrying IL6 597 GG genotype tended to develop aGVHD more frequently than IL6 597 recipients possessing at least one IL6 597*A allele


231

IL6 gene variants in subcohorts according to the type of aHSCT donor. To reflect type of aHSCT donor as one of the most important factors associated to the aHSCT immunobiology, we analysed the subcohorts of patients who were transplanted from related and unrelated donor separately. In related aHSCT subcohort, IL6 174 GG homozygous patients developed aGVHD in 17 out of 30 cases (56.7%) compared to 27 out of 86 patients with other genotypes (31.4%; P ¼ 0.01; OR ¼ 2.858; 95% CI 1.217–6.711; Figure 1). Accordingly, aGVHD was significantly more frequent in IL6 597 GG homozygous recipients than in those possessing at least one IL6 597*A allele (P ¼ 0.02). Furthermore, estimated frequency of IL6 174*G/597*G haplotype was higher in the group of patients who developed aGVHD (60.7%) compared to recipients without aGVHD (49.3%) but the difference was not significant (P ¼ 0.1). Analysing the association of IL6 gene variants with development of cGVHD we have not found any significant results. Survival analysis of related aHSCT pairs confirmed decreased OS in patients possessing at least one IL6 174*G allele (median survival time: 14.5 months) compared to IL6 174 CC homozygous recipients (median survival time: 36.5 months; log-rank test: P ¼ 0.04; Figure 2). A presence of IL6 174 GG genotype in recipient increased significantly TRM (mean survival time: 35.8 months; P ¼ 0.02) compared to recipients with IL6 174 GC and CC genotypes (mean survival time: 76.5 months; Figure 3a) and similar effect on TRM increase was apparent also for IL6 597 GG homozygous recipients compared to those with IL6 597 GA or AA genotypes (P ¼ 0.01; Figure 3b). In our limited group of 45 unrelated aHSCT pairs, investigated IL6 gene variants did not significantly affect on occurrence of acute or chronic GVHD nor modify survival after aHSCT. Particularly, IL6 174 GG homozygous recipients with an unrelated donor developed aGVHD in similar proportion (5/12 cases, 41.7%) as patients with other genotypes (40.0%; P ¼ NS; Figure 1). To complete our analysis we have examined the association of the donor’s IL6 gene variants with the occurrence of acute or chronic GVHD and influence of survival after aHSCT in univariate and multivariate

Transplant-related mortality

1.0

0.8

0.6

0.4

IL6 - 174 GG genotype

0.2 IL6 - 174 GC or CC genotype 0.0 0

2


10

12

1.0 Transplant-related mortality

(P ¼ 0.06). On the other hand, no relationship between IL6 gene polymorphisms and development of cGVHD was found. Apart from investigating the association of selected SNPs with GVHD, we were also interested in whether these polymorphisms are relevant for survival after HSCT and TRM. In patients possessing at least one IL6 174 G* allele, a borderline significance for decrease in OS was observed (median survival time: 14.6 months) compared to IL6 174 CC homozygous recipient (median survival time: 36.5 months; log-rank test: P ¼ 0.05). Furthermore, a significant increase in TRM was observed in IL6 597 GG homozygous recipients (mean survival time: 37.2 months) compared to patients with IL6 597 GC or CC genotypes (mean survival time: 71.8 months; P ¼ 0.04).

0.8

0.6

0.4

IL6 - 597 GG genotype

0.2 IL6 - 597 GA or AA genotype 0.0 0

2


10

12

Figure 3 Cumulative incidence of 1 year TRM in relation to IL6 174 and 597 genotypes in the aHSCT recipients transplanted from related donor. Cumulative mortality was significantly increased in IL6 174 (P ¼ 0.02, a) and 597 (P ¼ 0.01, b) GG patients compared to recipients with both other genotypes.

analysis of all cohorts. In our study, donor’s genotype did not affect any parameter of aHSCT clinical outcome (online Appendix 3).

Analysis of CCL2 gene polymorphisms association with aHSCT outcome CCL2 gene variants in the whole aHSCT group. Apart from investigating the IL6 SNPs, we were also interested in whether CCL2 2518 and 2076 gene polymorphisms are relevant for aHSCT clinical outcome. In the whole study group we have not found any association of these SNPs with the development of acute or chronic GVHD. Although no significant effect of CCL2 variants on OS was observed as well, the analysis of TRM revealed significant increase of mortality in CCL2 2076 TT homozygous patients (mean survival time 16.7 months) Bone Marrow Transplantation


232

compared to individuals carrying CCL2 2076*A allele (mean survival time 67.2 months; P ¼ 0.04). This difference in TRM was even stronger when CCL2 2076 TT homozygous patients were compared with those possessing ‘opposite’ homozygous genotype CCL2 2076 AA (mean survival time 70.2 months; P ¼ 0.02). CCL2 gene variants in subcohorts according to the type of aHSCT donor. The association of CCL2 2076 gene polymorphism with TRM after transplantation was seen also in the group of related aHSCT pairs; TRM was significantly increased in CCL2 2076 TT homozygous patients (mean survival time 15.0 months) compared to individuals carrying CCL2 2076 A*allele (mean survival time 70.9 months; P ¼ 0.02) or CCL2 2076 AA homozygous recipients (mean survival time 72.6 months; P ¼ 0.01). Furthermore, CCL2 2076 TT genotype in recipients appeared to associate with worse OS (median survival time 2.3 months) compared to recipients possessing CCL2 2076*A allele (median survival time 21.0 months; P ¼ 0.04) or CCL2 2076 AA homozygous individuals (median survival time 19.2 months; P ¼ 0.03). No association of CCL2 2518 and 2076 gene polymorphisms with the development of acute or chronic GVHD neither impact on OS or TRM was found in the group of patient who received their graft from the unrelated donor.

Multivariate analysis of aHSCT outcome To evaluate the role of known clinical and biological (genetic) factors considerable for the aHSCT outcome, multivariate regression analysis for the endpoints aGVHD and cGVHD was conducted, which, except of investigated IL6 and CCL2 SNP variants, included primary diagnosis, donor relationship, donor–recipient sex combination, conditioning regimen, stem-cell source and GVHD prophylaxis. In this model, only the combination of female donor to male recipient remained independent predictor of cGVHD in the subcohort of patients with related donor (P ¼ 0.03; OR 2.542; 95% CI 1.078–5.991). In case of OS, Cox regression model was adjusted for all above-mentioned covariates together with the presence of aGVHD and/or cGVHD in recipients. We revealed the presence of acute (P ¼ 1*106; OR 3.968; 95% CI 2.280– 6.907) and chronic (Po1*105; OR 5.943; 95% CI 2.742– 12.880) GVHD and the type of GVHD prophylaxis (P ¼ 0.03; OR 1.433; 95% CI 0.502–4.265) as independent factors associated with decrease of OS in the whole group. When the subcohort of patients with related donor has been analysed separately, the diagnosis of AML appeared to be further predictor of the OS after aHSCT (P ¼ 0.03; OR 2.031; 95% CI 1.060–3.890) in addition to aGVHD (Po1*105; OR 4.878; 95% CI 2.463–9.659) and chronic GVHD (P ¼ 0.001; OR 4.353; 95% CI 1.864–10.167). Nevertheless, the associations of IL6 and CCL2 variants with aHSCT outcome observed in univariate analyses have not been apparent at significant level in multivariate regression models (Table 3). Bone Marrow Transplantation

Table 3 Analysis of risk factors for OS for all aHSCT pairs and the groups of related and unrelated aHSCT Factor

All aHSCT

aGVHD P ¼ 0.000001 CGVHD Po0.00001 GVHD prophylaxis P ¼ 0.03 Diagnosis of AML NS

Related aHSCT Unrelated aHSCT Po0.00001 P ¼ 0.001 NS P ¼ 0.03

NS P ¼ 0.005 NS NS

Abbreviations: aGVHD ¼ acute GVHD; aHSCT ¼ allogeneic haematopoietic SCT; cGVHD ¼ chronic GVHD.

Discussion Today, aHSCT is a well-established treatment for lifethreatening haematological malignant and non-malignant diseases.1 However, in 30–50% of cases it is followed by various post-transplant complications which remain the chief cause of morbidity and mortality after aHSCT. Aiming at determination of new parameters for predicting aHSCT complications, a wide spectrum of non-HLA genes, including cytokines and chemokines, related to patients and donors has recently been investigated for their possible role in the development of GVHD, infections and their impact on TRM and OS.18–20 Proinflammatory cytokines and chemokines have been implicated in the pathogenesis of GVHD and their functional gene polymorphisms may affect each of its three phases.21 The first phase represents an effect of the conditioning regimen, which leads to the massive upregulation of proinflammatory cytokines and might be influenced by the patient genotype. In the second phase of GVHD, donor T-cell activation influenced by the donor and also recipient genotype, leads to release of proinflammatory cytokines (IL-1, IL-6, tumour necrosis factor-a) and migration of activated donor T cells potentiated by chemokines may cause pathological damage to targets organs such as skin, liver and gut. During the third phase, CTL and natural killer cells induce target tissue damage through cell-mediated cytotoxicity. This effect might be dependent on both patient and donor cytokine genotype.3,22 Given the participation of proinflammatory cytokines in the above-described mechanisms leading to GVHD, the first aim of our study was to investigate a possible association between development of post-transplantation complications and polymorphisms in the gene for a representative proinflammatory cytokine IL-6. Several investigations on the impact of IL6 gene polymorphism on aHSCT outcome have already been published but the results are somewhat inconsistent (Table 4). Some authors report an association between IL6 gene polymorphism on the side of the recipient and others, on the donor side. Some studies have found a significant impact of IL6 gene polymorphism on development of acute GVHD whereas in others particular IL6 gene variants are associated with increased incidence of chronic GVHD. A relationship between the carrying of the IL6 174*G allele of the recipient and development of acute GVHD was found by Cavet8 and Karabon.9 Association between acute GVHD and possession of IL6 174*G allele or presence of


233 Overview of the association studies of IL6 -174 G/C gene polymorphism with aHSCT outcome

Table 4 Author Socie´ et al.

4

Cavet et al.8

Karabon et al.9

Cohort

Studied association

Suggested association

Transplantation 2001

100 HLA-identical sibling aHSCT 80 HLA-identical sibling aHSCT

aGVHD, cGVHD

107 HLA-identical sibling aHSCT 160 HLA-matched related aHSCT 93 recipients, 74 donors—sibling aHSCT

Infection, aGVHD, cGVHD, OS, TRM aGVHD, cGVHD, OS

cGVHD and GG or GC genotype of the recipient (P ¼ 0.020) aGVHD and G allele in recipient (P ¼ 0.055), cGVHD and GG genotype of the recipient (P ¼ 0.005) and of the donor (P ¼ 0.061) No association

Blood 2001

Rocha et al.20 Mullighan et al.

Journal

Blood 2002 5

Transplantation 2004 Human Immunology 2005

aGVHD, cGVHD

aGVHD, cGVHD, OS

aGVHD and G allele in donor (P ¼ 0.001) aGVHD and G allele in recipient (P ¼ 0.078) and donor GG genotype (P ¼ 0.034)

Abbreviations: aGVHD ¼ acute GVHD; aHSCT ¼ allogeneic haematopoietic SCT; cGVHD ¼ chronic GVHD.

IL6 174 GG genotype in donor was first described by Mullighan5 and Karabon.9 In contrast, Socie´4 and Cavet8 described the contribution of IL6 174 GG or both GG and GC genotype of the recipient to the risk of development of chronic GVHD. In our study, we found that the presence of IL6 174 GG genotype in recipients was associated with the risk for development of acute GVHD. These recipients had also worse OS and increased incidence of TRM in the univariate analysis. Although we observed clear effect of IL6 174 GG genotype on deterioration of aHSCT outcome in the group of recipients transplanted from related donor, this association was completely absent among unrelated aHSCT pairs. Nevertheless, the multivariate analysis did not reveal IL6 genetic polymorphisms as the independent risk factors for the aHSCT clinical outcome. Above-mentioned discrepancy between the effect of IL6 174 SNP in our related and unrelated aHSCT pairs could generally be explained by several facts; firstly, there could be a role of a priori difference in the biology and clinical management of aHSCT in related and unrelated settings. Secondly, our group of unrelated aHSCT pairs consisted only of 45 patients and their respective donors which is, indeed, not sufficient for relevant conclusions in genetic association study. Nevertheless, our data combined with previous mostly positive reports on IL6 174*G allele or GG genotype in aHSCT (Table 4) allow the ‘mechanistic’ speculation that effect of these variants may be apparent only under particular clinical/biological conditions (for example, aHSCT from HLA-identical related donor in our situation). This explanation is in line with recent growing evidence that particular functional (regulatory) genetic elements play their role under specific conditions or in the presence of a specific stimulus.23,24 In the aHSCT situation, functionality and ‘penetrance’ of these variants may depend on interaction with other genes (for example, minor histocompatibility Ag) or environmental factors (for example, conditioning regimen, GVHD prophylaxis). As possible example, sharing of non-HLA polymorphic variants between sibling aHSCT pairs, which is by nature much higher than that in unrelated pairs, may facilitate specific gene–gene, or better ‘genome–genome’, interactions in aHSCT. However, these

speculations may only be confirmed in much larger groups of aHSCT pairs that enable to stratify reasonable numbers of particular ‘homogenous’ aHSCT complex phenotypes. From the hierarchical point of view and based on our multivariate analyses, the impact of IL6 gene variants on the aHSCT outcome may be considered rather as being ‘supplementary’ to main known clinical and biological parameters relevant to aHSCT outcome. In the second part of this study, we focused on analysing a possible association between the aHSCT outcome and gene polymorphism in another proinflammatory mediator–chemotactic cytokine MCP-1 (CCL2). One publication on solid organ transplantation reported that genetic determination of MCP-1 (CCL2) overproduction was associated with significant deterioration in kidney graft function.13 MCP-1 (CCL2) plays an important role in the migration of activated donor T cells to the target GVHD tissues.11 To the best of our knowledge, no study on CCL2 gene polymorphism and aHSCT outcome has been published to date. Comparison of CCL2 genotype frequencies revealed a significant lower frequency of CCL2 2076 AA genotype in patients than in the groups of their donors or healthy controls. This finding may suggest the possible genetic difference between patients and healthy individuals, which should be confirmed in another study. In our study, no significant association of CCL2 SNPs with development of acute or chronic GVHD was found. In survival analysis, only a trend towards shorter OS in the whole study cohort but its significant decrease in related aHSCT group in patients with CCL2 2076 TT genotype compared to CCL2 2076 AA homozygous individuals was observed. Similarly, analysis of TRM showed a significant increase in patients with CCL2 2076 TT genotype compared to CCL2 2076 AA homozygous individuals in the whole study group and also in recipients transplanted from related donor. Our finding of the possible impact of CCL2 2076 SNP for survival after aHSCT has to be confirmed in a larger cohort of recipients. At the same time, it must be determined whether it is the gene polymorphism itself that mediates the MCP-1 (CCL2) effect on the aHSCT outcome. Bone Marrow Transplantation


234

Study limitations The patients have been recruited in three clinical centres (Olomouc, Pilsen, Ljubljana). Therefore, heterogeneity in clinical management (for example, conditioning regimen, GVHD prophylaxis protocols) among centres may confound (mask) the effect of IL6 and CCL2 variants in specific clinical conditions. Nevertheless, the results of multivariate analyses controlling for the majority of considerable covariates are provided. No correction for number of investigated SNP markers has been applied, which increased the possibility to obtain false-positive data. Because the lack of previous reports on CCL2 gene variants in aHSCT, the results for this gene should be considered preliminary. Two further parameters which have been shown to interfere with the aHSCT outcome—gut decontamination25 and CD34 þ graft cell dose26—were not available for substantial proportion of patients, and, therefore, these could not be involved in multivariate analyses. In conclusion, from the univariate analysis, this investigation revealed an association of the SNP variants of the IL6 gene with GVHD, TRM and survival after aHSCT, which was apparent particularly in the related aHSCT setting. Further, preliminary evidence for a role of the investigated CCL2 gene polymorphisms on aHSCT outcome was found. In the context of previous findings, data from our study provides further support for the notion that IL6 genetic variants are important genetic factors, which can modify the aHSCT outcome under particular clinical conditions. We propose that metaanalysis of previous findings and improved selection of a larger set of aHSCT pairs should be performed to allow broader stratification of patients according to the relevant biological/clinical factors, which may identify aHSCT situations in which IL6 variants play a real role. Supplementary information is available at Association of IL6 and CCL2 gene polymorphisms with the outcome of allogeneic haematopoietic stem cell transplantation’s website (Appendices 1, 2 and 3).

Acknowledgements Technical assistance of S Zachova, M Lukesova, J Onderkova and A Stahelova is gratefully acknowledged. We thank I Kemperle for collection of clinical data of Slovenian aHSCT pairs. This study was supported by the Research Programme of the Ministry of Education, Youth and Sports No. 6198959205 and the Internal Grant Agency of the Ministry of Health No. NR9099.

References 1 Copelan EA. Hematopoietic stem-cell transplantation. N Engl J Med 2006; 354: 1813–1826. 2 Dickinson AM, Middleton PG, Rocha V, Gluckman E, Holler E, Eurobank members. Genetic polymorphisms predicting the outcome of bone marrow transplants. Br J Haematol 2004; 127: 479–490.


3 Dickinson AM, Charron D. Non-HLA immunogenetics in hematopoietic stem cell transplantation. Curr Opin Immunol 2005; 17: 517–525. 4 Socie` G, Loiseau P, Tamouza R, Janin A, Busson M, Gluckman E et al. Both genetic and clinical factors predict the development of graft-versus-host disease after allogeneic hematopoietic stem cell transplantation. Transplantation 2001; 72: 699–706. 5 Mullighan C, Heatley S, Doherty K, Szabo F, Grigg A, Hughes T et al. Non-HLA immunogenetic polymorphisms and the risk of complications after allogeneic hemopoietic stem-cell transplantation. Transplantation 2004; 27: 587–596. 6 Bertinetto FE, Dall’Omo AM, Mazzola GA, Rendine S, Berrino M, Bertola L et al. Role of non-HLA genetic polymorphisms in graft-versus-host disease after haematopoietic stem cell transplantation. Int J Immunogenet 2006; 33: 375–384. 7 Fishman D, Faulds G, Jeffrey R, Mohamed-Ali V, Yudkin JS, Humphries S et al. The effect of novel polymorphisms in the interleukin-6 (IL-6) gene on IL-6 transcription and plasma IL6 levels, and an association with systemic-onset juvenil chronic arthritis. J Clin Invest 1998; 102: 1369–1376. 8 Cavet J, Dickinson AM, Norden J, Taylor PR, Jackson GH, Middleton PG. Interferon-gamma and interleukin-6 gene polymorphisms associate with graft-versus-host disease in HLA-matched sibling bone marrow transplantation. Blood 2001; 98: 1594–1600. 9 Karabon L, Wysoczanska B, Bogunia-Kubik K, Suchnicki K, Lange A. IL-6 and IL-10 promoter gene polymorphisms of patients and donors of allogeneic sibling hematopoietic stem sell transplants associate with the risk of acute graft-versushost disease. Hum Immunol 2005; 66: 700–710. 10 Zlotnik A, Yoshie O, Nomiyama H. The chemokine and chemokine receptor superfamilies and their molecular evaluation. Genome Biol 2006; 7: 243. 11 Wysocki CA, Panoskaltsis-Mortari A, Blazar BR, Serody JS. Leukocyte migration and graft-versus-host disease. Blood 2005; 105: 4191–4199. 12 Rovin BH, Lu L, Saxena R. A novel polymorphism in the MCP-1 gene regulatory region that influences MCP-1 expression. Biochem Biophys Res Commun 1999; 259: 344–348. 13 Lacha J, Hribova P, Kotsch K, Brabcova I, Bartosova K, Volk HD et al. Effect of cytokines and chemokines (TGF-beta, TNF-alpha, IL-6, IL-10, MCP-1, RANTES) gene polymorphisms in kidney recipients on posttransplantation outcome: influence of donor-recipient match. Transplant Proc 2005; 37: 764–766. 14 Schro¨ppel B, Fischereder M, Lin M, Marder B, Schiano T, Kra¨mer BK et al. Analysis of gene polymorphisms in the regulatory region of MCP-1, RANTES and CCR5 in liver transplant recipients. J Clin Immunol 2002; 22: 381–385. 15 Kubistova Z, Mrazek F, Tudos Z, Kriegova E, Ambruzova Z, Mytilineos J et al. Distribution of 22 cytokine gene polymorphisms in the healthy Czech population. Int J Immunogenet 2006; 33: 261–267. 16 Bunce M, O’Neill CM, Barnardo MC, Krausa P, Browning MJ, Morris PJ et al. Phototyping: comprehensive DNA typing for HLA-A, B, C, DRB3, DRB4, DRB5 & DQB1 by PCR with 144 primer mixes utilizing sequence-specific primers (PCR-SSP). Tissue Antigens 1995; 46: 355–367. 17 Excoffier L, Laval G, Schneider S. Arlequin ver. 3.0: an integrated software package for population genetics data analysis. Evol Bioinform Online 2005; 1: 47–50. 18 Dickinson AM, Middleton PG. Beyond the HLA typing age: genetic polymorphisms predicting transplant outcome. Blood Rev 2005; 19: 333–340.


235 19 Dickinson AM. Risk assessment in haematopoietic stem cell transplantation: pre-transplant patient and donor factors: non-HLA genetics. Best Pract Res Clin Haematol 2007; 20: 189–207. 20 Rocha V, Franco RF, Porcher R, Bittencourt H, Silva Jr WA, Latouche A et al. Host defense and inflammatory gene polymorphisms are associated with outcomes after HLA-identical sibling bone marrow transplantation. Blood 2002; 100: 3908–3918. 21 Jaksch M, Mattsson J. The pathophysiology of acute graftversus-host disease. Scand J Immunol 2005; 61: 398–409. 22 Holler E. Cytokines, viruses, and graft-versus-host disease. Curr Opin Hematol 2002; 9: 479–484. 23 Kallberg H, Padyukov L, Plenge RM, Ronnelid J, Gregersen PK, van der Helm-van Mil AH et al. Epidemiological Investigation of Rheumatoid Arthritis study group. gene-gene and gene-environment interactions involving HLA-DRB1,

PTPN22, and smoking in two subsets of rheumatoid arthritis. Am J Hum Genet 2007; 80: 867–875. 24 Huttunen R, Aittoniemi J, Laine J, Vuento R, Karjalainen J, Rovio AT et al. Gene-environment interaction between MBL2 genotype and smoking, and the risk of gram-positive bacteraemia. Scand J Immunol 2008; 68: 438–444. 25 Holler E, Rogler G, Brenmoehl J, Hahn J, Herfarth H, Greinix H et al. Prognostic significance of NOD2/CARD15 variants in HLA-identical sibling hematopoietic stem cell transplantation: effect on long-term outcome is confirmed in 2 independent cohorts and may be modulated by the type of gastrointestinal decontamination. Blood 2006; 107: 4189–4193. 26 Dı´ ez-Campelo M, Pe´rez-Simoń JA, Ocio EM, Castilla C, Gonza´lez-Porras JR, Sańchez-Guijo FM et al. CD34+ cell dose and outcome of patients undergoing reduced-intensityconditioning allogeneic peripheral blood stem cell transplantation. Leuk Lymphoma 2005; 46: 177–183.

Supplementary Information accompanies the paper on Bone Marrow Transplantation website (http://www.nature.com/bmt)
