- Email: [email protected]
The associations between interleukin-1 polymorphisms and susceptibility to ankylosing spondylitis: A meta-analysis
Joint Bone Spine 79 (2012) 370–374
Original article
The associations between interleukin-1 polymorphisms and susceptibility to ankylosing spondylitis: A meta-analysis Wu-in Lea a , Young Ho Lee b,∗ a
Korea University College of Medicine, 26-1, Anam-dong 5-ga, Seongbuk-gu, Seoul 136-705, Korea Division of Rheumatology, Department of Internal Medicine, Korea University Anam Hospital, Korea University College of Medicine, 26-1, Anam-dong 5-ga, Seongbuk-gu, Seoul 136-705, Korea b
a r t i c l e
i n f o
Article history: Accepted 23 June 2011 Available online 1 October 2011 Keywords: Interleukin-1 Polymorphism Ankylosing spondylitis Meta-analysis
a b s t r a c t Objective: The aim of this study was to determine whether polymorphisms of interleukin-1 (IL-1) confer susceptibility to ankylosing spondylitis (AS). Methods: The authors conducted meta-analyses on associations between IL-1 polymorphisms and AS susceptibility, using fixed or random effects models. In order to avoid duplications and data previously subjected to meta-analysis, we performed meta-analysis on studies if new data on IL-1 polymorphisms in AS were reported. Results: A total of nine studies consisting of 20 separate comparisons of association between IL-1 polymorphisms and AS susceptibility were included in this meta-analysis. These were performed on European, Asian, and Latin American population samples. Meta-analysis revealed a significant association between the 2 allele of the IL-1F10.3 polymorphism (rs3811581) and the risk of developing AS in Europeans (OR = 0.775, 95% CI = 0.605–0.992, P = 0.043). Furthermore, the OR of the 2 allele of IL-1A + 889 (rs1800587) was found to be significantly increased in Europeans with AS (OR = 1.357, 95% CI = 1.085–1.697, P = 0.007). However, meta-analyses of the IL-1B-511, IL-1B + 3953, and ILF7.1 polymorphisms and of the variable numbers of tandem repeats of the IL-1 receptor antagonist gene (IL-1RN VNTR) revealed no association between AS and these polymorphisms. Conclusions: In addition to the three known IL-1 polymorphisms, rs2856836, rs17561, and rs1894399, found in previous meta-analysis, this meta-analysis shows that the IL-1F10.3 and IL-1A + 889 polymorphisms are associated with the development of AS in Europeans but not in Asians. © 2011 Société franc¸aise de rhumatologie. Published by Elsevier Masson SAS. All rights reserved.
1. Introduction Ankylosing spondylitis (AS) is a chronic inflammatory disorder characterized by inflammation in the spinal and sacroiliac joints, which initially causes bone and joint erosion, and eventually leads to ankylosis [1]. Strong genetic factors have been implicated in the etiology of this disease. Human leukocyte antigen (HLA) B27 is the first genetic factor identified in AS, and has been shown to confer greatest susceptibility to AS [2]. However, there is increasing evidence that non-HLA-B27 genes also contribute to AS susceptibility [1]. The chromosome 2q13 is a linkage region of AS. The interleukin1 (IL-1) family gene cluster lies 132 cM from the p-telomere of chromosome 2 in a region linked to AS susceptibility [3]. IL-1 is secreted by activated macrophages in inflamed synovium, and initiates the recruitment of immune cells and inflammation. The IL-1
∗ Corresponding author. Tel.: +822 920 5645; fax: +822 922 5974. E-mail address: [email protected] (Y.H. Lee).
gene family contains nine genes, IL-1A, IL-1B, IL-1F7, IL-1F9, IL1F6, IL-1F8, IL-1F5, IL-1F10, and IL-1 receptor antagonist (IL-1RN) genes in a cluster within an ∼360-kb region [4]. Six of these genes, IL-1F7, IL-1F9, IL-1F6, IL-1F8, IL-1F5, and IL-1F10, share structural homology with IL-1A/B (proinflammatory IL-1 agonists) and IL1RN (anti-inflammatory IL-1 antagonist) and lie between IL-1A and IL-1RN [5]. Previous studies have revealed that IL-1 polymorphisms, including two single nucleotide polymorphisms (SNPs) at positions −511 and +3954 of the IL-1B gene, and a variable number of tandem repeats of the IL-1 receptor antagonist gene (IL-1RN VNTR) are associated with an increased risk of autoimmune disease development [6]. IL-1 polymorphisms have been reported to be associated with AS susceptibility in some, but not in all studies [7–15]. These result discrepancies may be due to small sample sizes, low statistical power, or clinical heterogeneity. In the present study, we adopted a meta-analysis approach to overcome the limitations of individual studies and resolve these disparities [16–18]. The aim of the present study was to determine whether IL-1 polymorphisms confer susceptibility to AS.
1297-319X/$ – see front matter © 2011 Société franc¸aise de rhumatologie. Published by Elsevier Masson SAS. All rights reserved. doi:10.1016/j.jbspin.2011.06.010
W.-i. Lea, Y.H. Lee / Joint Bone Spine 79 (2012) 370–374
2. Methods 2.1. Identification of eligible studies and data extraction We performed a search for studies that examined associations between IL-1 polymorphisms and AS. A search of the literature was conducted using MEDLINE to identify available articles (published up to July 2010) in which IL-1 polymorphisms were determined in AS patients and controls. Combinations of keywords, such as, ‘interleukin-1’, ‘IL-1’, ‘polymorphism’, ‘ankylosing spondylitis’, and ‘AS’, were entered as medical subject heading (MeSH) and text words. References in identified studies were also investigated to identify additional studies not indexed by MEDLINE. Genetic association studies that determined the distributions of IL-1 genotypes in AS cases and controls were eligible for inclusion. Studies were included in the analysis if; (1) they were case-control studies, (2) they contained original data, (3) they provided enough data to calculate odds ratios (ORs). The following information was extracted from each study; first author, year of publication, racial makeup of the study population, numbers of cases and controls, and genotype and allele frequency information for IL-1 polymorphisms. 2.2. Evaluation of publication bias Funnel plots are used to detect publication bias, but they require a range of studies of varying sizes and subjective judgments, and thus, we evaluated publication bias using Egger’s linear regression test [19], which measures funnel plot asymmetry on a natural logarithm scale of ORs. 2.3. Evaluation of statistical associations Allelic frequencies of the IL-1 polymorphisms from the respective studies were determined by the allele counting method. We examined the contrast of the allelic effect of 2 (minor allele) vs. 1 (common allele) of the IL-1 polymorphisms. And for IL-1RN VNTR, allele 1 (IL-1RN*1), containing four repeats, was the most common. The remaining alleles, representing three (IL-1RN*4), five (IL-1RN*3) and six repeats (IL-1RN*5) of a tandem of 86 bp length occur rarely in most general populations. The IL-1RN*1/*3/*4/*5 alleles are all categorized as IL-1RN long (L alleles) [20]. Two repeats (IL-1RN*2) has been reported to be associated with autoimmune diseases [21]. We performed meta-analysis of 2 allele vs. L allele. Point estimates of risk, ORs, and their 95% confidence intervals (CIs) were estimated for each study. We assessed within- and between-study variations or heterogeneity using Cochran’s Q-statistic [22]. This heterogeneity test assesses the null hypothesis that all the studies are evaluating the same effect. We also quantified the effect of heterogeneity by using I2 = 100% × (Q – df)/Q [23]. I2 ranges between 0 and 100% and represents the proportion of between-study variability that can be attributed to heterogeneity rather than chance. I2 values of 25, 50, and 75% were defined as low, moderate, and high estimates, respectively. Fixed effects assumes that genetic factors have similar effects on AS susceptibility across all investigated studies, and that observed variations between studies are caused by chance alone [24]. The random effects model assumes that different studies show substantial diversity and assesses both within-study sampling errors and between-study variances [25]. If study groups show no heterogeneity, the fixed and random effects models produce similar results, and if not the random effects model usually produces wider CIs than the fixed effects model. The random effects model is used when between-study heterogeneity is evident [24]. Statistical manipulations were undertaken using a comprehensive meta-analysis computer program (Biosta, Englewood, NJ, USA). Previous meta-analysis has been performed on the relations between IL-1 polymorphisms at AS [8]. In order to
371
avoid duplicated and update data used in previous meta-analysis, we performed meta-analysis on studies if new data on IL-1 polymorphisms in AS were reported. We performed meta-analysis of associations between polymorphisms and AS, if there were at least two comparisons. 3. Results 3.1. Studies included in the meta-analysis Electronic and manual searches resulted in the identification of 18 studies, and 13 were selected for full-text review based on their titles and abstracts [7–15,26–29]. Four studies were excluded because they contained no extractable data [26,27], or family linkage data [28,29]. A total of nine relevant studies met the study inclusion criteria [7–15]. One of the eligible studies contained data on 10 different groups [8] and another study contained data on three different groups [10]. These groups were treated independently, and hence a total of 20 separate comparisons were available for analysis. The genotype data from the Canadian and Taiwanese cohorts in Sims et al.’s study [8] had previously been reported by Chou et al. [11] and by Maksymowych et al. [10], and the data on rs419598, rs315952, and the IL-1RN VNTR used in UK cohort of Sims et al.’s study [8] have previously been published in Timms et al. [12]. Thus, we included only one data to avoid data duplication. Due to the limited number of studies on the polymorphisms, we performed six types of meta-analyses on the IL-1B-511 (rs16944), IL-1B + 3953 (rs1143634), IL-1A + 889 (rs1800587), IL-1F10.3 (rs3811581), IL-1F7.1 (rs2723187) and IL1RN VNTR polymorphisms. The selected characteristics of the nine studies determining the relationship between the IL-1 polymorphisms and the risks of AS are summarized in Table 1. 3.2. Meta-analysis of the IL-1B-511 and IL-1B + 3953 polymorphisms and ankylosing spondylitis susceptibility Meta-analysis failed to identify an association between AS and the IL-1B-511 2 allele in the total population (OR = 0.918, 95% CI = 0.812–1.038, P = 0.173) (Table 2, Fig. 1). Furthermore, stratification by ethnicity failed to identify any association between this polymorphism and AS in the European, Asian, or Latin American groups (Table 2, Fig. 1). In addition, no association was observed between AS susceptibility and the IL-1B + 3953 polymorphism by meta-analyses. 3.3. Meta-analysis of the IL-1F10.3, IL-1F7.1, IL-1A + 889 polymorphisms and ankylosing spondylitis susceptibility Meta-analysis revealed a significant association between the 2 allele of the IL-1F10.3 polymorphism and the risk of developing AS in Europeans (OR = 0.775, 95% CI = 0.605–0.992, P = 0.043) (Table 2, Fig. 2). However, in Asians and Latin Americans, no associational difference was found (Fig. 2). Meta-analysis revealed no association between AS and the IL-1F7.1 polymorphism in the European population (OR = 1.273, 95% CI = 0.879–1.845, P = 0.201) (Fig. 1). However, the OR of the 2 allele of IL-1A + 889 was found to be significantly increased in Europeans (OR = 1.357, 95% CI = 1.085–1.697, P = 0.007) (Fig. 3). 3.4. Meta-analysis of the association between IL-1RN VNTR and ankylosing spondylitis No association was found between AS susceptibility and IL-1RN VNTR by meta-analysis (OR = 1.071, 95% CI = 0.897–1.278, P = 0.450) (Table 2, Fig. 1). Furthermore, stratification by ethnicity failed to
372
W.-i. Lea, Y.H. Lee / Joint Bone Spine 79 (2012) 370–374
Table 1 Characteristics of individual studies included in the meta-analysis. Study [Ref]
Country (ethnicity)
NumbersAS/control
IL-1 polymorphism(s)
Major findings for association
Guo et al., 2010 [7]
China (A)
240/240
Sims et al., 2008 [8]
Australia (M)
2675/2592
IL-1F10.3 (association P-value = 0.001), B-511-T/B + 3953-C/F10.3-C/RN4T/RN2VNTR-1/RN6.1-C (P = 3.32 × 10−5 ) rs2856836, rs17561, rs1894399(P = 0.0036, 0.000019, 0.0003)
Agrawal et al., 2008 [9] Maksymowych et al., 2006 [10]
India (E) Canada (E)
162/111 394/446
Chou et al., 2006 [11]
China (A)
200/200
Timms et al., 2004 [12]
UK (E)
529/200
Maksymowych et al., 2003 [13]
Canada (E)
394/500
van der Paardt et al., 2002 [14] McGarry et al., 2001 [15]
Netherlands (E) UK (E)
106/104 188/155
IL-1RN VNTR, IL-1B-511 (rs16944), IL-1B + 3953 (rs1143634), IL-1F10.3 (rs3811058), IL-1RN4 (rs419598), IL-1RN6/1 (rs31952) IL-1A (rs2856836, rs17561, rs1894399), IL-1B (rs16944), IL-1F10 (rs3811058), IL-1RA (rs419598, rs315952, rs315951), IL-1RN VNTR IL-1RN VNTR rs2856836, rs3783550, rs3783547, rs3783543, rs17561, rs3783526, rs1800794, rs1143643, rs1143634, rs1143630, rs3917354, rs1143627, rs3811047, rs2723187, rs895497, rs1900287, rs3811058, rs419598, rs315951 IL-1B-511 (rs16944), IL-1F10.3 (rs3811058), IL-1RN4 (rs419598), IL-1RN6/1 (rs315952), IL-1RN6//2 (rs315951), IL-1RN VNTR IL-1A559, A376, A889, IL-1B3953, B5810, B-511, IL-1F7-1, F7-2, LOCUS 129359, IL-1F9-1, F6-1, F6-2, F6-3, F8-1, F8-2, F8-3, F5-1, F5-2, F5-3, F5-4, F10-1, F10-2, F10-3, IL-1RN8006, RN8061, RN9589, RN11100, IL-1RN VNTR IL-1RN exon 4 (rs27810), exon 6 (rs30735, rs31017) IL-1RN VNTR, IL-1B + 3953, IL-1B-511 IL-1A-889, IL-1B-511, IL-1RN VNTR
IL-1RN VNTR (P = 0.0002) rs3783526 (IL-1A) (P = 0.0009), rs1143627 (IL-1B) (P = 0.0005), rs1143634/rs1143630/rs3917356 and rs1143630/rs3917356/rs3917354 (P = 0.006-0.0001)
IL-1F10.3, IL-1RN4, IL-1RN VNTR (P < 0.05), IL-1RN4/IL-1RN VNTR (P = 0.004) IL-1A559 (P = 0.003), A889 (P = 0.025), IL-1B5810 (P = 0.023), B-511 (P = 0.0038), IL-1F6-1 (P = 0.037), F6-3 (P = 0.0005), F8-2 (P = 0.039), F5-1 (P = 0.045), F5-2 (P = 0.0002), F5-3 (P = 0.004) rs30735 (P = 0.001), rs31017 (P = 0.04) IL-1RN VNTR (P = 0.031) IL-1RN VNTR (P = 0.001)
AS: ankylosing spondylitis; Ref: reference; E: European; A: Asian; LA: Latin American; M: multiple ethnicities; UK: United Kingdom; NS: not significant; IL-1RN VNTR: a variable number of tandem repeats of the IL-1 receptor antagonist gene.
identify any association between this polymorphism and AS in Europeans, Asians, or Latin Americans (Table 2, Fig. 1).
analysis was small. Egger’s regression test showed no evidence of publication bias in this meta-analysis of IL-1 polymorphisms in all groups (Egger’s regression test P-values > 0.1).
3.5. Heterogeneity and publication bias 4. Discussion Some between-study heterogeneity was found during the metaanalyses, but no evidence of heterogeneity was found for analyses of IL-1A + 899 polymorphisms in the European population (Table 2). It was difficult to correlate the funnel plot, which is usually used to detect publication bias, as the number of studies included in the
In the present study, we combined evidence of associations between IL-1 polymorphisms and AS susceptibility. The results of this meta-analysis provide evidence of an association between the IL-1 polymorphisms and AS. Although our findings do not support
Table 2 Results of meta-analysis on the associations between IL-1 polymorphisms and ankylosing spondylitis. Polymorphism
IL1RN 2 vs. L allele
IL-1B-511 2 vs. 1 allele
IL-1B + 3953 2 vs. 1 allele IL-1F10.3 2 vs. 1 allele
IL-1F7.1 2 vs. 1 allele IL-1A + 889 2 vs. 1 allele
Population
Overall European Asian Latin American Overall European Asian Latin American Overall European Asian Overall European Asian Latin American Overall European Overall European
No. of studies
13 10 2 1 15 11 3 1 6 5 1 13 9 3 1 4 4 2 2
Numbers
Test of association
Test of heterogeneity
Patients
Control
OR
95% CI
P-val
Model
P-val
I2
4782 3384 788 86 5363 4641 624 98 2046 1566 480 5075 3959 1018 98 1430 1430 1094 1094
4394 3110 694 90 4383 3553 738 92 1482 1030 452 4396 3172 1132 92 778 778 564 564
1.071 1.130 0.685 1.355 0.918 0.911 0.953 1.355 1.130 1.103 2.292 0.805 0.775 0.905 0.640 1.273 1.273 1.357 1.357
0.897–1.278 0.940–1.360 0.373–1.257 0.683–2.687 0.812-1.038 0.777–1.070 0.802–1.132 0.683–2.687 0.369–1.364 0.911–1.336 0.801–6.559 0.647–1.002 0.605–0.992 0.542–1.512 0. 300–1.367 0.879–1.845 0.879–1.845 1.085–1.697 1.085–1.697
0.450 0.194 0.222 0.385 0.173 0.256 0.582 0.385 0.205 0.314 0.122 0.052 0.043 0.704 0.249 0.201 0.201 0.007 0.007
R R R NA R R F NA F F NA R R R NA F F F F
0.002 0.006 0.080 NA 0.016 0.005 0.493 NA 0.301 0.372 NA 0.001 0.084 0.000 NA 0.788 0.788 0.828 0.828
62.0 60.9 67.2 NA 49.3 60.3 0 NA 17.4 6.02 NA 62.8 42.4 88.0 NA 0 0 0 0
AS: ankylosing spondylitis; IL-1RN VNTR: variable number of tandem repeats of the IL-1 receptor antagonist gene; L: any long allele (1, 3, 4, or 5); F: fixed effects model; R: random effects model; NA: not available.
W.-i. Lea, Y.H. Lee / Joint Bone Spine 79 (2012) 370–374
373
Fig. 3. Odds ratios and 95% CIs of individual studies and pooled data for the 2 vs. 1 allele of the IL-1A + 889 polymorphism with respect to susceptibility to ankylosing spondylitis in Europeans.
Fig. 1. Odds ratios and 95% CIs of individual studies and pooled data for the 2 vs. 1 allele of the IL1RN (A), IL-1B-511 (B), IL-1B + 3953 (C), and IL-1F7.1 (D) polymorphism for susceptibility to ankylosing spondylitis in overall group.
Fig. 2. Odds ratios and 95% CIs of individual studies and pooled data for the 2 vs. 1 allele of the IL-1F10.3 polymorphism for susceptibility to ankylosing spondylitis in Europeans, Asians, and Latin Americans.
associations between the IL-1B-511, IL-1B + 3953, ILF7.1, and the IL-1RN VNTR polymorphisms and AS susceptibility, meta-analysis revealed a significant association between the IL-1F10.3 polymorphism and the risk of developing AS in Europeans, but not in Asians and Latin Americans. Furthermore, our meta-analysis showed an association between AS and the IL-1A + 889 polymorphism in Europeans. These findings suggest that the IL-1F10.3 and the IL-1A + 899 polymorphisms are associated with the development of AS in Europeans. Our analysis differs slightly from the previous meta-analysis performed on relations between the IL-1 polymorphisms and AS performed by Sims et al. [8]. In this previous study, nine polymorphisms in the IL-1 gene cluster members were analyzed, namely, IL-1A (rs2856836, rs17561 and rs1894399), IL-1B (rs16944), IL-1F10.3 (rs3811058) and IL-1RN (rs419598, the IL1RN VNTR, rs315952, and rs315951) in a large sample, which included 2675 AS cases and 2592 controls, and an association was found between three polymorphisms (rs2856836, rs17561, and rs1894399) and AS. In contrast, we updated the previous metaanalysis performed by Sims et al. [8] and combined all available data on the IL-1 polymorphisms in AS, and performed meta-analyses on the relations between the IL-1B-511 (rs16944), IL-1B + 3953 (rs1143634), IL-1A + 889 (rs180587), IL-1F10.3 (rs3811058), IL-1F7.1 (rs2723187), and IL-1RN VNTR polymorphisms and AS. There were no new data about three polymorphisms (rs2856836, rs17561, and rs1894399) since previous meta-analysis [8]. In previous metaanalyses, subgroup analysis was not performed by ethnicity [8], but we subjected published data to meta-analysis to evaluate the genetic associations between IL-1 polymorphisms and AS susceptibility in different populations. In particular, the IL-1F10.3 (rs3811058) polymorphism was not found to be associated with AS by the meta-analysis conducted by Sims et al. [8]. Our updated meta-analysis, which included two more studies [7,9], also found no such association in all population, but ethnicity-specific analysis revealed a significant association between the polymorphism and AS in Europeans, but not in Asians. And our meta-analysis revealed that the IL-1A + 889 polymorphism was significantly associated in Europeans. The IL-1A + 889 polymorphism was not studied in the previous meta-analysis [8]. Thus, this meta-analysis added that the IL-1F10.3 and IL-1A + 889 polymorphisms were associated with AS in Europeans, in addition to three IL-1 polymorphisms, namely, rs2856836, rs17561, and rs1894399, identified in previous meta-analyses. The present study has some limitations that require consideration. First, our study could suffer from publication bias, heterogeneity, or from the effects of confounding factors. Analysis of data stratified by gender, disease activity, severity, or HLA-B27 status would have provided more information, but this information was not available from the studies examined. Second, ethnicspecific meta-analysis included data from European and Asian patients, and thus, our results are only applicable to only these ethnic groups. Third, most of the studies were performed in populations of European descent, and only a handful of studies were conducted in Asian and Latin American populations. Thus, further
374
W.-i. Lea, Y.H. Lee / Joint Bone Spine 79 (2012) 370–374
studies are needed in other ethnic populations. Fourth, the possibility that IL-1 haplotypes contribute to the development of AS also needs to be examined by meta-analysis [13]. In the present study, we could not conduct meta-analysis on IL-1 susceptibility haplotypes due to limited data. Fifth, IL-1 polymorphisms have been reported to be associated with the severity rather than susceptibility to autoimmune diseases [30], but we could not analyze the associations between IL-1 polymorphisms and the severity of AS due to limited data. In conclusion, in addition to three IL-1 polymorphisms, namely, rs2856836, rs17561, and rs1894399, identified in previous metaanalyses, this meta-analysis of published data finds that the IL-1F10.3 and IL-1A + 889 polymorphisms are associated with AS in Europeans. Disclosure of interest The authors declare that they have no conflicts of interest concerning this article. References [1] Brown MA, Wordsworth BP, Reveille JD. Genetics of ankylosing spondylitis. Clin Exp Rheumatol 2002;20:S43–9. [2] Brown MA, Laval SH, Brophy S, et al. Recurrence risk modelling of the genetic susceptibility to ankylosing spondylitis. Ann Rheum Dis 2000;59:883–6. [3] Laval SH, Timms A, Edwards S, et al. Whole-genome screening in ankylosing spondylitis: evidence of non-MHC genetic-susceptibility loci. Am J Hum Genet 2001;68:918–26. [4] Laurincova B. Interleukin-1 family: from genes to human disease. Acta Univ Palacki Olomuc Fac Med 2000;143:19–29. [5] Smith DE, Renshaw BR, Ketchem RR, et al. Four new members expand the interleukin-1 superfamily. J Biol Chem 2000;275:1169–75. [6] Lee YH, Ji JD, Song GG. Association between interleukin 1 polymorphisms and rheumatoid arthritis susceptibility: a meta-analysis. J Rheumatol 2009;36:12–5. [7] Guo ZS, Li C, Lin ZM, et al. Association of IL-1 gene complex members with ankylosing spondylitis in Chinese Han population. Int J Immunogenet 2010;37:33–7. [8] Sims AM, Timms AE, Bruges-Armas J, et al. Prospective meta-analysis of interleukin 1 gene complex polymorphisms confirms associations with ankylosing spondylitis. Ann Rheum Dis 2008;67:1305–9. [9] Agrawal S, Srivastava R, Sharma B, et al. IL1RN*2 allele of IL-1receptor antagonist VNTR polymorphism is associated with susceptibility to ankylosing [corrected] spondylitis in Indian patients. Clin Rheumatol 2008;27:573–6. [10] Maksymowych WP, Rahman P, Reeve JP, et al. Association of the IL1 gene cluster with susceptibility to ankylosing spondylitis: an analysis of three Canadian populations. Arthritis Rheum 2006;54:974–85.
[11] Chou CT, Timms AE, Wei JC, et al. Replication of association of IL1 gene complex members with ankylosing spondylitis in Taiwanese Chinese. Ann Rheum Dis 2006;65:1106–9. [12] Timms AE, Crane AM, Sims AM, et al. The interleukin 1 gene cluster contains a major susceptibility locus for ankylosing spondylitis. Am J Hum Genet 2004;75:587–95. [13] Maksymowych WP, Reeve JP, Reveille JD, et al. High-throughput singlenucleotide polymorphism analysis of the IL1RN locus in patients with ankylosing spondylitis by matrix-assisted laser desorption ionization-time-offlight mass spectrometry. Arthritis Rheum 2003;48:2011–8. [14] van der Paardt M, Crusius JB, Garcia-Gonzalez MA, et al. Interleukin-1beta and interleukin-1 receptor antagonist gene polymorphisms in ankylosing spondylitis. Rheumatology (Oxford) 2002;41:1419–23. [15] McGarry F, Neilly J, Anderson N, et al. A polymorphism within the interleukin 1 receptor antagonist (IL-1Ra) gene is associated with ankylosing spondylitis. Rheumatology (Oxford) 2001;40:1359–64. [16] Lee YH, Rho YH, Choi SJ, et al. The PTPN22 C1858 T functional polymorphism and autoimmune diseases–a meta-analysis. Rheumatology (Oxford) 2007;46:49–56. [17] Lee YH, Witte T, Momot T, et al. The mannose-binding lectin gene polymorphisms and systemic lupus erythematosus: two case-control studies and a meta-analysis. Arthritis Rheum 2005;52:3966–74. [18] Lee YH, Harley JB, Nath SK. Meta-analysis of TNF-alpha promoter -308 A/G polymorphism and SLE susceptibility. Eur J Hum Genet 2006;14:364–71. [19] Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34. [20] Tarlow JK, Blakemore AI, Lennard A, et al. Polymorphism in human IL-1 receptor antagonist gene intron 2 is caused by variable numbers of an 86-bp tandem repeat. Hum Genet 1993;91:403–4. [21] Arman A, Yilmaz B, Coker A, et al. Interleukin-1 receptor antagonist (IL-1RN) and interleukin-1B gene polymorphisms in Turkish patients with rheumatoid arthritis. Clin Exp Rheumatol 2006;24:643–8. [22] Davey Smith G, Egger M. Meta-analyses of randomised controlled trials. Lancet 1997;350:1182. [23] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. [24] Egger M, Smith GD, Phillips AN. Meta-analysis: principles and procedures. BMJ 1997;315:1533–7. [25] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88. [26] Kim TJ, Kim TH, Lee HJ, et al. Interleukin 1 polymorphisms in patients with ankylosing spondylitis in Korea. J Rheumatol 2008;35:1603–8. [27] Kim TH, Stone MA, Rahman P, et al. Interleukin 1 and nuclear factor-kappaB polymorphisms in ankylosing spondylitis in Canada and Korea. J Rheumatol 2005;32:1907–10. [28] Jin L, Zhang G, Akey JM, et al. Lack of linkage of IL1RN genotypes with ankylosing spondylitis susceptibility. Arthritis Rheum 2004;50:3047–8. [29] Djouadi K, Nedelec B, Tamouza R, et al. Interleukin 1 gene cluster polymorphisms in multiplex families with spondylarthropathies. Cytokine 2001;13:98–103. [30] Buchs N, di Giovine FS, Silvestri T, et al. IL-1B and IL-1Ra gene polymorphisms and disease severity in rheumatoid arthritis: interaction with their plasma levels. Genes Immun 2001;2:222–8.
Original article
The associations between interleukin-1 polymorphisms and susceptibility to ankylosing spondylitis: A meta-analysis Wu-in Lea a , Young Ho Lee b,∗ a
Korea University College of Medicine, 26-1, Anam-dong 5-ga, Seongbuk-gu, Seoul 136-705, Korea Division of Rheumatology, Department of Internal Medicine, Korea University Anam Hospital, Korea University College of Medicine, 26-1, Anam-dong 5-ga, Seongbuk-gu, Seoul 136-705, Korea b
a r t i c l e
i n f o
Article history: Accepted 23 June 2011 Available online 1 October 2011 Keywords: Interleukin-1 Polymorphism Ankylosing spondylitis Meta-analysis
a b s t r a c t Objective: The aim of this study was to determine whether polymorphisms of interleukin-1 (IL-1) confer susceptibility to ankylosing spondylitis (AS). Methods: The authors conducted meta-analyses on associations between IL-1 polymorphisms and AS susceptibility, using fixed or random effects models. In order to avoid duplications and data previously subjected to meta-analysis, we performed meta-analysis on studies if new data on IL-1 polymorphisms in AS were reported. Results: A total of nine studies consisting of 20 separate comparisons of association between IL-1 polymorphisms and AS susceptibility were included in this meta-analysis. These were performed on European, Asian, and Latin American population samples. Meta-analysis revealed a significant association between the 2 allele of the IL-1F10.3 polymorphism (rs3811581) and the risk of developing AS in Europeans (OR = 0.775, 95% CI = 0.605–0.992, P = 0.043). Furthermore, the OR of the 2 allele of IL-1A + 889 (rs1800587) was found to be significantly increased in Europeans with AS (OR = 1.357, 95% CI = 1.085–1.697, P = 0.007). However, meta-analyses of the IL-1B-511, IL-1B + 3953, and ILF7.1 polymorphisms and of the variable numbers of tandem repeats of the IL-1 receptor antagonist gene (IL-1RN VNTR) revealed no association between AS and these polymorphisms. Conclusions: In addition to the three known IL-1 polymorphisms, rs2856836, rs17561, and rs1894399, found in previous meta-analysis, this meta-analysis shows that the IL-1F10.3 and IL-1A + 889 polymorphisms are associated with the development of AS in Europeans but not in Asians. © 2011 Société franc¸aise de rhumatologie. Published by Elsevier Masson SAS. All rights reserved.
1. Introduction Ankylosing spondylitis (AS) is a chronic inflammatory disorder characterized by inflammation in the spinal and sacroiliac joints, which initially causes bone and joint erosion, and eventually leads to ankylosis [1]. Strong genetic factors have been implicated in the etiology of this disease. Human leukocyte antigen (HLA) B27 is the first genetic factor identified in AS, and has been shown to confer greatest susceptibility to AS [2]. However, there is increasing evidence that non-HLA-B27 genes also contribute to AS susceptibility [1]. The chromosome 2q13 is a linkage region of AS. The interleukin1 (IL-1) family gene cluster lies 132 cM from the p-telomere of chromosome 2 in a region linked to AS susceptibility [3]. IL-1 is secreted by activated macrophages in inflamed synovium, and initiates the recruitment of immune cells and inflammation. The IL-1
∗ Corresponding author. Tel.: +822 920 5645; fax: +822 922 5974. E-mail address: [email protected] (Y.H. Lee).
gene family contains nine genes, IL-1A, IL-1B, IL-1F7, IL-1F9, IL1F6, IL-1F8, IL-1F5, IL-1F10, and IL-1 receptor antagonist (IL-1RN) genes in a cluster within an ∼360-kb region [4]. Six of these genes, IL-1F7, IL-1F9, IL-1F6, IL-1F8, IL-1F5, and IL-1F10, share structural homology with IL-1A/B (proinflammatory IL-1 agonists) and IL1RN (anti-inflammatory IL-1 antagonist) and lie between IL-1A and IL-1RN [5]. Previous studies have revealed that IL-1 polymorphisms, including two single nucleotide polymorphisms (SNPs) at positions −511 and +3954 of the IL-1B gene, and a variable number of tandem repeats of the IL-1 receptor antagonist gene (IL-1RN VNTR) are associated with an increased risk of autoimmune disease development [6]. IL-1 polymorphisms have been reported to be associated with AS susceptibility in some, but not in all studies [7–15]. These result discrepancies may be due to small sample sizes, low statistical power, or clinical heterogeneity. In the present study, we adopted a meta-analysis approach to overcome the limitations of individual studies and resolve these disparities [16–18]. The aim of the present study was to determine whether IL-1 polymorphisms confer susceptibility to AS.
1297-319X/$ – see front matter © 2011 Société franc¸aise de rhumatologie. Published by Elsevier Masson SAS. All rights reserved. doi:10.1016/j.jbspin.2011.06.010
W.-i. Lea, Y.H. Lee / Joint Bone Spine 79 (2012) 370–374
2. Methods 2.1. Identification of eligible studies and data extraction We performed a search for studies that examined associations between IL-1 polymorphisms and AS. A search of the literature was conducted using MEDLINE to identify available articles (published up to July 2010) in which IL-1 polymorphisms were determined in AS patients and controls. Combinations of keywords, such as, ‘interleukin-1’, ‘IL-1’, ‘polymorphism’, ‘ankylosing spondylitis’, and ‘AS’, were entered as medical subject heading (MeSH) and text words. References in identified studies were also investigated to identify additional studies not indexed by MEDLINE. Genetic association studies that determined the distributions of IL-1 genotypes in AS cases and controls were eligible for inclusion. Studies were included in the analysis if; (1) they were case-control studies, (2) they contained original data, (3) they provided enough data to calculate odds ratios (ORs). The following information was extracted from each study; first author, year of publication, racial makeup of the study population, numbers of cases and controls, and genotype and allele frequency information for IL-1 polymorphisms. 2.2. Evaluation of publication bias Funnel plots are used to detect publication bias, but they require a range of studies of varying sizes and subjective judgments, and thus, we evaluated publication bias using Egger’s linear regression test [19], which measures funnel plot asymmetry on a natural logarithm scale of ORs. 2.3. Evaluation of statistical associations Allelic frequencies of the IL-1 polymorphisms from the respective studies were determined by the allele counting method. We examined the contrast of the allelic effect of 2 (minor allele) vs. 1 (common allele) of the IL-1 polymorphisms. And for IL-1RN VNTR, allele 1 (IL-1RN*1), containing four repeats, was the most common. The remaining alleles, representing three (IL-1RN*4), five (IL-1RN*3) and six repeats (IL-1RN*5) of a tandem of 86 bp length occur rarely in most general populations. The IL-1RN*1/*3/*4/*5 alleles are all categorized as IL-1RN long (L alleles) [20]. Two repeats (IL-1RN*2) has been reported to be associated with autoimmune diseases [21]. We performed meta-analysis of 2 allele vs. L allele. Point estimates of risk, ORs, and their 95% confidence intervals (CIs) were estimated for each study. We assessed within- and between-study variations or heterogeneity using Cochran’s Q-statistic [22]. This heterogeneity test assesses the null hypothesis that all the studies are evaluating the same effect. We also quantified the effect of heterogeneity by using I2 = 100% × (Q – df)/Q [23]. I2 ranges between 0 and 100% and represents the proportion of between-study variability that can be attributed to heterogeneity rather than chance. I2 values of 25, 50, and 75% were defined as low, moderate, and high estimates, respectively. Fixed effects assumes that genetic factors have similar effects on AS susceptibility across all investigated studies, and that observed variations between studies are caused by chance alone [24]. The random effects model assumes that different studies show substantial diversity and assesses both within-study sampling errors and between-study variances [25]. If study groups show no heterogeneity, the fixed and random effects models produce similar results, and if not the random effects model usually produces wider CIs than the fixed effects model. The random effects model is used when between-study heterogeneity is evident [24]. Statistical manipulations were undertaken using a comprehensive meta-analysis computer program (Biosta, Englewood, NJ, USA). Previous meta-analysis has been performed on the relations between IL-1 polymorphisms at AS [8]. In order to
371
avoid duplicated and update data used in previous meta-analysis, we performed meta-analysis on studies if new data on IL-1 polymorphisms in AS were reported. We performed meta-analysis of associations between polymorphisms and AS, if there were at least two comparisons. 3. Results 3.1. Studies included in the meta-analysis Electronic and manual searches resulted in the identification of 18 studies, and 13 were selected for full-text review based on their titles and abstracts [7–15,26–29]. Four studies were excluded because they contained no extractable data [26,27], or family linkage data [28,29]. A total of nine relevant studies met the study inclusion criteria [7–15]. One of the eligible studies contained data on 10 different groups [8] and another study contained data on three different groups [10]. These groups were treated independently, and hence a total of 20 separate comparisons were available for analysis. The genotype data from the Canadian and Taiwanese cohorts in Sims et al.’s study [8] had previously been reported by Chou et al. [11] and by Maksymowych et al. [10], and the data on rs419598, rs315952, and the IL-1RN VNTR used in UK cohort of Sims et al.’s study [8] have previously been published in Timms et al. [12]. Thus, we included only one data to avoid data duplication. Due to the limited number of studies on the polymorphisms, we performed six types of meta-analyses on the IL-1B-511 (rs16944), IL-1B + 3953 (rs1143634), IL-1A + 889 (rs1800587), IL-1F10.3 (rs3811581), IL-1F7.1 (rs2723187) and IL1RN VNTR polymorphisms. The selected characteristics of the nine studies determining the relationship between the IL-1 polymorphisms and the risks of AS are summarized in Table 1. 3.2. Meta-analysis of the IL-1B-511 and IL-1B + 3953 polymorphisms and ankylosing spondylitis susceptibility Meta-analysis failed to identify an association between AS and the IL-1B-511 2 allele in the total population (OR = 0.918, 95% CI = 0.812–1.038, P = 0.173) (Table 2, Fig. 1). Furthermore, stratification by ethnicity failed to identify any association between this polymorphism and AS in the European, Asian, or Latin American groups (Table 2, Fig. 1). In addition, no association was observed between AS susceptibility and the IL-1B + 3953 polymorphism by meta-analyses. 3.3. Meta-analysis of the IL-1F10.3, IL-1F7.1, IL-1A + 889 polymorphisms and ankylosing spondylitis susceptibility Meta-analysis revealed a significant association between the 2 allele of the IL-1F10.3 polymorphism and the risk of developing AS in Europeans (OR = 0.775, 95% CI = 0.605–0.992, P = 0.043) (Table 2, Fig. 2). However, in Asians and Latin Americans, no associational difference was found (Fig. 2). Meta-analysis revealed no association between AS and the IL-1F7.1 polymorphism in the European population (OR = 1.273, 95% CI = 0.879–1.845, P = 0.201) (Fig. 1). However, the OR of the 2 allele of IL-1A + 889 was found to be significantly increased in Europeans (OR = 1.357, 95% CI = 1.085–1.697, P = 0.007) (Fig. 3). 3.4. Meta-analysis of the association between IL-1RN VNTR and ankylosing spondylitis No association was found between AS susceptibility and IL-1RN VNTR by meta-analysis (OR = 1.071, 95% CI = 0.897–1.278, P = 0.450) (Table 2, Fig. 1). Furthermore, stratification by ethnicity failed to
372
W.-i. Lea, Y.H. Lee / Joint Bone Spine 79 (2012) 370–374
Table 1 Characteristics of individual studies included in the meta-analysis. Study [Ref]
Country (ethnicity)
NumbersAS/control
IL-1 polymorphism(s)
Major findings for association
Guo et al., 2010 [7]
China (A)
240/240
Sims et al., 2008 [8]
Australia (M)
2675/2592
IL-1F10.3 (association P-value = 0.001), B-511-T/B + 3953-C/F10.3-C/RN4T/RN2VNTR-1/RN6.1-C (P = 3.32 × 10−5 ) rs2856836, rs17561, rs1894399(P = 0.0036, 0.000019, 0.0003)
Agrawal et al., 2008 [9] Maksymowych et al., 2006 [10]
India (E) Canada (E)
162/111 394/446
Chou et al., 2006 [11]
China (A)
200/200
Timms et al., 2004 [12]
UK (E)
529/200
Maksymowych et al., 2003 [13]
Canada (E)
394/500
van der Paardt et al., 2002 [14] McGarry et al., 2001 [15]
Netherlands (E) UK (E)
106/104 188/155
IL-1RN VNTR, IL-1B-511 (rs16944), IL-1B + 3953 (rs1143634), IL-1F10.3 (rs3811058), IL-1RN4 (rs419598), IL-1RN6/1 (rs31952) IL-1A (rs2856836, rs17561, rs1894399), IL-1B (rs16944), IL-1F10 (rs3811058), IL-1RA (rs419598, rs315952, rs315951), IL-1RN VNTR IL-1RN VNTR rs2856836, rs3783550, rs3783547, rs3783543, rs17561, rs3783526, rs1800794, rs1143643, rs1143634, rs1143630, rs3917354, rs1143627, rs3811047, rs2723187, rs895497, rs1900287, rs3811058, rs419598, rs315951 IL-1B-511 (rs16944), IL-1F10.3 (rs3811058), IL-1RN4 (rs419598), IL-1RN6/1 (rs315952), IL-1RN6//2 (rs315951), IL-1RN VNTR IL-1A559, A376, A889, IL-1B3953, B5810, B-511, IL-1F7-1, F7-2, LOCUS 129359, IL-1F9-1, F6-1, F6-2, F6-3, F8-1, F8-2, F8-3, F5-1, F5-2, F5-3, F5-4, F10-1, F10-2, F10-3, IL-1RN8006, RN8061, RN9589, RN11100, IL-1RN VNTR IL-1RN exon 4 (rs27810), exon 6 (rs30735, rs31017) IL-1RN VNTR, IL-1B + 3953, IL-1B-511 IL-1A-889, IL-1B-511, IL-1RN VNTR
IL-1RN VNTR (P = 0.0002) rs3783526 (IL-1A) (P = 0.0009), rs1143627 (IL-1B) (P = 0.0005), rs1143634/rs1143630/rs3917356 and rs1143630/rs3917356/rs3917354 (P = 0.006-0.0001)
IL-1F10.3, IL-1RN4, IL-1RN VNTR (P < 0.05), IL-1RN4/IL-1RN VNTR (P = 0.004) IL-1A559 (P = 0.003), A889 (P = 0.025), IL-1B5810 (P = 0.023), B-511 (P = 0.0038), IL-1F6-1 (P = 0.037), F6-3 (P = 0.0005), F8-2 (P = 0.039), F5-1 (P = 0.045), F5-2 (P = 0.0002), F5-3 (P = 0.004) rs30735 (P = 0.001), rs31017 (P = 0.04) IL-1RN VNTR (P = 0.031) IL-1RN VNTR (P = 0.001)
AS: ankylosing spondylitis; Ref: reference; E: European; A: Asian; LA: Latin American; M: multiple ethnicities; UK: United Kingdom; NS: not significant; IL-1RN VNTR: a variable number of tandem repeats of the IL-1 receptor antagonist gene.
identify any association between this polymorphism and AS in Europeans, Asians, or Latin Americans (Table 2, Fig. 1).
analysis was small. Egger’s regression test showed no evidence of publication bias in this meta-analysis of IL-1 polymorphisms in all groups (Egger’s regression test P-values > 0.1).
3.5. Heterogeneity and publication bias 4. Discussion Some between-study heterogeneity was found during the metaanalyses, but no evidence of heterogeneity was found for analyses of IL-1A + 899 polymorphisms in the European population (Table 2). It was difficult to correlate the funnel plot, which is usually used to detect publication bias, as the number of studies included in the
In the present study, we combined evidence of associations between IL-1 polymorphisms and AS susceptibility. The results of this meta-analysis provide evidence of an association between the IL-1 polymorphisms and AS. Although our findings do not support
Table 2 Results of meta-analysis on the associations between IL-1 polymorphisms and ankylosing spondylitis. Polymorphism
IL1RN 2 vs. L allele
IL-1B-511 2 vs. 1 allele
IL-1B + 3953 2 vs. 1 allele IL-1F10.3 2 vs. 1 allele
IL-1F7.1 2 vs. 1 allele IL-1A + 889 2 vs. 1 allele
Population
Overall European Asian Latin American Overall European Asian Latin American Overall European Asian Overall European Asian Latin American Overall European Overall European
No. of studies
13 10 2 1 15 11 3 1 6 5 1 13 9 3 1 4 4 2 2
Numbers
Test of association
Test of heterogeneity
Patients
Control
OR
95% CI
P-val
Model
P-val
I2
4782 3384 788 86 5363 4641 624 98 2046 1566 480 5075 3959 1018 98 1430 1430 1094 1094
4394 3110 694 90 4383 3553 738 92 1482 1030 452 4396 3172 1132 92 778 778 564 564
1.071 1.130 0.685 1.355 0.918 0.911 0.953 1.355 1.130 1.103 2.292 0.805 0.775 0.905 0.640 1.273 1.273 1.357 1.357
0.897–1.278 0.940–1.360 0.373–1.257 0.683–2.687 0.812-1.038 0.777–1.070 0.802–1.132 0.683–2.687 0.369–1.364 0.911–1.336 0.801–6.559 0.647–1.002 0.605–0.992 0.542–1.512 0. 300–1.367 0.879–1.845 0.879–1.845 1.085–1.697 1.085–1.697
0.450 0.194 0.222 0.385 0.173 0.256 0.582 0.385 0.205 0.314 0.122 0.052 0.043 0.704 0.249 0.201 0.201 0.007 0.007
R R R NA R R F NA F F NA R R R NA F F F F
0.002 0.006 0.080 NA 0.016 0.005 0.493 NA 0.301 0.372 NA 0.001 0.084 0.000 NA 0.788 0.788 0.828 0.828
62.0 60.9 67.2 NA 49.3 60.3 0 NA 17.4 6.02 NA 62.8 42.4 88.0 NA 0 0 0 0
AS: ankylosing spondylitis; IL-1RN VNTR: variable number of tandem repeats of the IL-1 receptor antagonist gene; L: any long allele (1, 3, 4, or 5); F: fixed effects model; R: random effects model; NA: not available.
W.-i. Lea, Y.H. Lee / Joint Bone Spine 79 (2012) 370–374
373
Fig. 3. Odds ratios and 95% CIs of individual studies and pooled data for the 2 vs. 1 allele of the IL-1A + 889 polymorphism with respect to susceptibility to ankylosing spondylitis in Europeans.
Fig. 1. Odds ratios and 95% CIs of individual studies and pooled data for the 2 vs. 1 allele of the IL1RN (A), IL-1B-511 (B), IL-1B + 3953 (C), and IL-1F7.1 (D) polymorphism for susceptibility to ankylosing spondylitis in overall group.
Fig. 2. Odds ratios and 95% CIs of individual studies and pooled data for the 2 vs. 1 allele of the IL-1F10.3 polymorphism for susceptibility to ankylosing spondylitis in Europeans, Asians, and Latin Americans.
associations between the IL-1B-511, IL-1B + 3953, ILF7.1, and the IL-1RN VNTR polymorphisms and AS susceptibility, meta-analysis revealed a significant association between the IL-1F10.3 polymorphism and the risk of developing AS in Europeans, but not in Asians and Latin Americans. Furthermore, our meta-analysis showed an association between AS and the IL-1A + 889 polymorphism in Europeans. These findings suggest that the IL-1F10.3 and the IL-1A + 899 polymorphisms are associated with the development of AS in Europeans. Our analysis differs slightly from the previous meta-analysis performed on relations between the IL-1 polymorphisms and AS performed by Sims et al. [8]. In this previous study, nine polymorphisms in the IL-1 gene cluster members were analyzed, namely, IL-1A (rs2856836, rs17561 and rs1894399), IL-1B (rs16944), IL-1F10.3 (rs3811058) and IL-1RN (rs419598, the IL1RN VNTR, rs315952, and rs315951) in a large sample, which included 2675 AS cases and 2592 controls, and an association was found between three polymorphisms (rs2856836, rs17561, and rs1894399) and AS. In contrast, we updated the previous metaanalysis performed by Sims et al. [8] and combined all available data on the IL-1 polymorphisms in AS, and performed meta-analyses on the relations between the IL-1B-511 (rs16944), IL-1B + 3953 (rs1143634), IL-1A + 889 (rs180587), IL-1F10.3 (rs3811058), IL-1F7.1 (rs2723187), and IL-1RN VNTR polymorphisms and AS. There were no new data about three polymorphisms (rs2856836, rs17561, and rs1894399) since previous meta-analysis [8]. In previous metaanalyses, subgroup analysis was not performed by ethnicity [8], but we subjected published data to meta-analysis to evaluate the genetic associations between IL-1 polymorphisms and AS susceptibility in different populations. In particular, the IL-1F10.3 (rs3811058) polymorphism was not found to be associated with AS by the meta-analysis conducted by Sims et al. [8]. Our updated meta-analysis, which included two more studies [7,9], also found no such association in all population, but ethnicity-specific analysis revealed a significant association between the polymorphism and AS in Europeans, but not in Asians. And our meta-analysis revealed that the IL-1A + 889 polymorphism was significantly associated in Europeans. The IL-1A + 889 polymorphism was not studied in the previous meta-analysis [8]. Thus, this meta-analysis added that the IL-1F10.3 and IL-1A + 889 polymorphisms were associated with AS in Europeans, in addition to three IL-1 polymorphisms, namely, rs2856836, rs17561, and rs1894399, identified in previous meta-analyses. The present study has some limitations that require consideration. First, our study could suffer from publication bias, heterogeneity, or from the effects of confounding factors. Analysis of data stratified by gender, disease activity, severity, or HLA-B27 status would have provided more information, but this information was not available from the studies examined. Second, ethnicspecific meta-analysis included data from European and Asian patients, and thus, our results are only applicable to only these ethnic groups. Third, most of the studies were performed in populations of European descent, and only a handful of studies were conducted in Asian and Latin American populations. Thus, further
374
W.-i. Lea, Y.H. Lee / Joint Bone Spine 79 (2012) 370–374
studies are needed in other ethnic populations. Fourth, the possibility that IL-1 haplotypes contribute to the development of AS also needs to be examined by meta-analysis [13]. In the present study, we could not conduct meta-analysis on IL-1 susceptibility haplotypes due to limited data. Fifth, IL-1 polymorphisms have been reported to be associated with the severity rather than susceptibility to autoimmune diseases [30], but we could not analyze the associations between IL-1 polymorphisms and the severity of AS due to limited data. In conclusion, in addition to three IL-1 polymorphisms, namely, rs2856836, rs17561, and rs1894399, identified in previous metaanalyses, this meta-analysis of published data finds that the IL-1F10.3 and IL-1A + 889 polymorphisms are associated with AS in Europeans. Disclosure of interest The authors declare that they have no conflicts of interest concerning this article. References [1] Brown MA, Wordsworth BP, Reveille JD. Genetics of ankylosing spondylitis. Clin Exp Rheumatol 2002;20:S43–9. [2] Brown MA, Laval SH, Brophy S, et al. Recurrence risk modelling of the genetic susceptibility to ankylosing spondylitis. Ann Rheum Dis 2000;59:883–6. [3] Laval SH, Timms A, Edwards S, et al. Whole-genome screening in ankylosing spondylitis: evidence of non-MHC genetic-susceptibility loci. Am J Hum Genet 2001;68:918–26. [4] Laurincova B. Interleukin-1 family: from genes to human disease. Acta Univ Palacki Olomuc Fac Med 2000;143:19–29. [5] Smith DE, Renshaw BR, Ketchem RR, et al. Four new members expand the interleukin-1 superfamily. J Biol Chem 2000;275:1169–75. [6] Lee YH, Ji JD, Song GG. Association between interleukin 1 polymorphisms and rheumatoid arthritis susceptibility: a meta-analysis. J Rheumatol 2009;36:12–5. [7] Guo ZS, Li C, Lin ZM, et al. Association of IL-1 gene complex members with ankylosing spondylitis in Chinese Han population. Int J Immunogenet 2010;37:33–7. [8] Sims AM, Timms AE, Bruges-Armas J, et al. Prospective meta-analysis of interleukin 1 gene complex polymorphisms confirms associations with ankylosing spondylitis. Ann Rheum Dis 2008;67:1305–9. [9] Agrawal S, Srivastava R, Sharma B, et al. IL1RN*2 allele of IL-1receptor antagonist VNTR polymorphism is associated with susceptibility to ankylosing [corrected] spondylitis in Indian patients. Clin Rheumatol 2008;27:573–6. [10] Maksymowych WP, Rahman P, Reeve JP, et al. Association of the IL1 gene cluster with susceptibility to ankylosing spondylitis: an analysis of three Canadian populations. Arthritis Rheum 2006;54:974–85.
[11] Chou CT, Timms AE, Wei JC, et al. Replication of association of IL1 gene complex members with ankylosing spondylitis in Taiwanese Chinese. Ann Rheum Dis 2006;65:1106–9. [12] Timms AE, Crane AM, Sims AM, et al. The interleukin 1 gene cluster contains a major susceptibility locus for ankylosing spondylitis. Am J Hum Genet 2004;75:587–95. [13] Maksymowych WP, Reeve JP, Reveille JD, et al. High-throughput singlenucleotide polymorphism analysis of the IL1RN locus in patients with ankylosing spondylitis by matrix-assisted laser desorption ionization-time-offlight mass spectrometry. Arthritis Rheum 2003;48:2011–8. [14] van der Paardt M, Crusius JB, Garcia-Gonzalez MA, et al. Interleukin-1beta and interleukin-1 receptor antagonist gene polymorphisms in ankylosing spondylitis. Rheumatology (Oxford) 2002;41:1419–23. [15] McGarry F, Neilly J, Anderson N, et al. A polymorphism within the interleukin 1 receptor antagonist (IL-1Ra) gene is associated with ankylosing spondylitis. Rheumatology (Oxford) 2001;40:1359–64. [16] Lee YH, Rho YH, Choi SJ, et al. The PTPN22 C1858 T functional polymorphism and autoimmune diseases–a meta-analysis. Rheumatology (Oxford) 2007;46:49–56. [17] Lee YH, Witte T, Momot T, et al. The mannose-binding lectin gene polymorphisms and systemic lupus erythematosus: two case-control studies and a meta-analysis. Arthritis Rheum 2005;52:3966–74. [18] Lee YH, Harley JB, Nath SK. Meta-analysis of TNF-alpha promoter -308 A/G polymorphism and SLE susceptibility. Eur J Hum Genet 2006;14:364–71. [19] Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34. [20] Tarlow JK, Blakemore AI, Lennard A, et al. Polymorphism in human IL-1 receptor antagonist gene intron 2 is caused by variable numbers of an 86-bp tandem repeat. Hum Genet 1993;91:403–4. [21] Arman A, Yilmaz B, Coker A, et al. Interleukin-1 receptor antagonist (IL-1RN) and interleukin-1B gene polymorphisms in Turkish patients with rheumatoid arthritis. Clin Exp Rheumatol 2006;24:643–8. [22] Davey Smith G, Egger M. Meta-analyses of randomised controlled trials. Lancet 1997;350:1182. [23] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. [24] Egger M, Smith GD, Phillips AN. Meta-analysis: principles and procedures. BMJ 1997;315:1533–7. [25] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88. [26] Kim TJ, Kim TH, Lee HJ, et al. Interleukin 1 polymorphisms in patients with ankylosing spondylitis in Korea. J Rheumatol 2008;35:1603–8. [27] Kim TH, Stone MA, Rahman P, et al. Interleukin 1 and nuclear factor-kappaB polymorphisms in ankylosing spondylitis in Canada and Korea. J Rheumatol 2005;32:1907–10. [28] Jin L, Zhang G, Akey JM, et al. Lack of linkage of IL1RN genotypes with ankylosing spondylitis susceptibility. Arthritis Rheum 2004;50:3047–8. [29] Djouadi K, Nedelec B, Tamouza R, et al. Interleukin 1 gene cluster polymorphisms in multiplex families with spondylarthropathies. Cytokine 2001;13:98–103. [30] Buchs N, di Giovine FS, Silvestri T, et al. IL-1B and IL-1Ra gene polymorphisms and disease severity in rheumatoid arthritis: interaction with their plasma levels. Genes Immun 2001;2:222–8.