CED
Clinical dermatology • Original article
Clinical and Experimental Dermatology
Association of the tumour necrosis factor-a polymorphisms rs361525 and rs1800629 with susceptibility to psoriasis: a meta-analysis Y. Jia,1 H. J. Qin,2 J. X. Zhang,1 X. L. Liu1 and L. J. Li1 1
First Affiliated Hospital and 2School of Forensic Medicine, Henan University of Science and Technology, Luoyang, China
doi:10.1111/ced.12136
Summary
Background. Evidence has suggested that tumour necrosis factor (TNF)-a may be involved in the aetiology of psoriasis, but the underlying association of the TNF-a polymorphisms 238G/A (rs361525) and 308G/A (rs1800629) with the risk of psoriasis is still unconfirmed. Aim. This meta-analysis was performed to determine whether the TNF-a 238G/A and 308G/A polymorphisms are associated with susceptibility to psoriasis. Methods. Eligible studies were identified by searching PubMed, EMBASE, CNKI (China National Knowledge Infrastructure), CBM (Chinese biomedical literature database) and WANFANG databases within a range of published years from 1990 to August 2012. The odds ratio (OR) and 95% confidence interval (CI) were used to assess the different associations. Results. In total, 17 studies with 2847 cases and 2222 controls were found for 238G/A and 20 studies with 2975 cases and 2243 controls for 308G/A. The pooled results showed an overall increased risk of psoriasis for the 238G/A polymorphism (OR = 2.06, 95% CI = 1.45–2.94, P < 0.001 for AA/GA vs. GG) and a reduced psoriasis risk with the 308G/A polymorphism (OR = 0.68, 95% CI = 0.59–0.79, P < 0.001 for AA/GA vs. GG). This association was only present in early-onset psoriasis (OR = 3.68, 95% CI = 2.17–6.24, P < 0.001 for 238G/A; OR = 0.56, 95% CI = 0.43–0.72, P < 0.001 for 308G/A), whereas there was no association (OR = 0.98, 95% CI = 0.56–1.70, P = 0.92 for 238G/A) or a unreliable association (OR = 0.66, 95% CI = 0.46–0.94, P = 0.02 for 308G/A) in late-onset psoriasis. Conclusions. This meta-analysis suggests that the TNF-a –238 and –308 promoter polymorphisms may play different roles in conferring susceptibility to psoriasis. Functional and well-designed studies should be conducted to confirm these results.
Introduction Psoriasis is a common, chronic inflammatory skin disease, affecting 2–3% of the global population. Two Correspondence: Dr Lijun Li, First Affiliated Hospital, Henan University of Science and Technology, Luoyang 471003, China E-mail: [email protected] Conflict of interest: none declared. The first two authors contributed equally to this work and should be considered joint first authors. Accepted for publication 16 December 2012
836
Clinical and Experimental Dermatology (2013) 38, pp836–844
types are recognised, based on age of onset. Type I (early-onset) psoriasis has an age of onset of < 40 years, and tends to be familial and severe. Type II (late-onset) occurs after the age of 40 years, and is more likely to be sporadic.1 Psoriasis is also divided based on clinical symptoms into psoriasis vulgaris (PV), pustular psoriasis, erythrodermic psoriasis, and psoriatic arthritis (PsA). Approximately 5–30% of patients with psoriasis may also develop PsA. Psoriasis is a T-cell-mediated disease, driven at least partly by a positive feedback loop from activated T cells to antigenpresenting cells, which is mediated by interferon-c,
ª 2013 British Association of Dermatologists
TNF-a polymorphism and psoriasis Y. Jia et al.
interleukin-1 and tumour necrosis factor (TNF)-a.1 TNF-a is a proinflammatory cytokine that has been implicated in the pathogenesis of a large number of human diseases.2 TNF-a levels are increased in psoriatic lesions, and biological drugs that block TNF-a are highly effective treatments for psoriasis, strongly supporting the role for TNF-a in the disease pathogenesis.1 Genome-wide association studies have identified several psoriasis susceptibility loci, including a locus on the short arm of chromosome at 6p21 (PSORS1).3 The TNF-a gene is also located nearby (6p21.3).2,4 Several polymorphisms have been identified in the TNF-a promoter region. The single nucleotide polymorphisms 238G/A (rs361525) and 308G/A (rs1800629) have been shown to directly affect TNF-a expression, and are therefore candidates for genetic psoriasis research.2,5 Several studies have explored the potential association between TNF-a promoter polymorphisms and susceptibility to psoriasis, but the results were inconsistent.3,6–11 A meta-analysis was performed in 2007, but it focused only on PV and had a relatively small sample size.4 Hence, we conducted a metaanalysis to investigate the proposed correlation more thoroughly.
Methods Literature search strategy
Electronic searches were conducted of PubMed, EMBASE, CNKI (China National Knowledge Infrastructure), CBM (Chinese biomedical literature database) and WANFANG databases (updated to August 6, 2012), using the terms ‘psoriasis’, ‘psoriatic arthritis’, ‘tumour necrosis factoralpha’ or ‘TNF-alpha’ and ‘polymorphism’. No restrictions were placed on language. Only published journal articles and academic dissertations were included. The reference lists of the research papers were also searched to identify other relevant publications. Inclusion and exclusion criteria
All studies included in the meta-analysis were required to meet the following criteria: (i) case–control study; (ii) evaluating the association of TNF-a rs361525 (-238G/A) and/or rs1800629 (-308G/A) polymorphisms with psoriasis risk; and (iii) presenting sufficient available data to assess odds ratio (OR) and 95% CI. Studies with insufficient or duplicate data were excluded. For overlapping and republished studies, the first study published or the one with the largest sample size (depending on the data) were included.
ª 2013 British Association of Dermatologists
Data extraction
All data were extracted independently by two reviewers (JY and QH) according to the prespecified selection criteria, and were reviewed and checked by a third investigator (ZJ). Disagreements about inclusion of studies and interpretation of data were resolved by discussion. Statistical analysis
Hardy–Weinberg equilibrium (HWE) was assessed in control groups using the Fisher exact test, with R software (version 2.15.0; R Foundation for Statistical Computing, Vienna, Austria) before statistical analysis. Dichotomous data are presented as OR with 95% CI. Statistical heterogeneity was measured using the Qstatistic and I2 test. A fixed-effects model was used to estimate the summary OR when there was no heterogeneity; otherwise, the random effects model was used. To explore sources of heterogeneity across studies, logistic meta-regression analyses were conducted. Sensitivity analyses were performed to investigate the influence of individual studies on the summary-effect estimates. The Begg rank correlation method and Egger weighted regression method were also used to assess the publication bias. The statistical analyses were performed using Stata software (version 12.0; StataCorp, College Station, TX, USA), and P < 0.05 was considered significant.
Results Study characteristics
The search strategy retrieved 69 potentially relevant articles. Based on the inclusion criteria, only 20 articles,3,5–23 consisting of 21 studies (1 article was considered to have 2 studies with 2 different populations) were included in the meta-analysis. Of these 21 studies, 17 studied TNF-a 238G/A in 2847 cases and 2222 controls, and 20 studies assessed TNF-a 308G/A in 2975 cases and 2243 controls. The corresponding characteristics are shown in Tables 1 and 2. Meta-analysis of
238G/A
The frequencies of the A allele of 238G/A ranged from 0.01 to 0.17 in control subjects. Because the AA genotype was not included in some studies, the pooled examination was carried out on the dominant genetic model (AA/GA vs. GG). This showed that carriage of the A allele was associated with an increased risk of
Clinical and Experimental Dermatology (2013) 38, pp836–844
837
TNF-a polymorphism and psoriasis Y. Jia et al.
Table 1 Characteristics of the included studies on the tumour necrosis factor-a
238G/A (rs361525) polymorphism. Genotypes n
Author
Year
Region
Ethnicity
Hohler6
1997
Germany
White
Hamamoto13 Gonzalez15 AI-Heresh16 Li17
2000 2001 2002 2002
Japan Spain UK China
East Asian White White East Asian
Reich5
2002
Germany
White
Zhang18
2002
China
East Asian
Kim19 Tsunemi7 Long8
2003 2003 2004
Korea Japan China
East Asian East Asian East Asian
Mossner9 Rahman10 Rahman10 Nedoszytko21
2005 2006 2006 2007
Germany Canada (N) Canada (Toronto) Poland
White White White White
Reich3
2007
Germany
White
Magalhaes23 Gallo11
2010 2012
Brazil Spain
White White
Classification of psoriasis P type PsA PsA PsA PsA P type P type P type P type P type P type PV PV P type P type PV PsA PsA P type P type PV PsA P type PV
I (PV)*
I II I (PV) II (PV) I (PV) II (PV)
I (PV) II (PV)
I (PV) II (PV)
I
Samples, n
Cases
Cases
Controls
GG
GA
AA
GG
GA
AA
PHWE
60 62 20 52 124 77 46 156 75 65 22 103 163 48 29 239 237 203 134 32 375 375 69 81
99 99 87 73 101 165 165 345 345 128 128 125 96 82 82 135 103 101 65 65 376 376 70 71
37 42 20 45 101 56 44 114 64 48 20 89 158 36 28 182 181 164 94 31 273 303 48 70
18 20 0 7 23 19 1 41 11 17 2 14 5 9 1 55 45 35 35 1 97 60 21 10
5 0 0 0 0 2 1 1 0 0 0 0 0 3 0 2 2 0 5 0 3 5 0 1
92 92 83 70 88 142 142 316 316 114 114 111 94 79 79 123 93 90 62 62 338 338 51 50
7 7 4 3 12 20 20 29 29 12 12 13 2 3 3 12 9 10 2 2 32 32 19 18
0 0 0 0 1 3 3 0 0 2 2 1 0 0 0 0 1 1 1 1 2 2 0 3
1.00 1.00 1.00 1.00 0.77 0.12 0.12 1.00 1.00 0.14 0.14 0.72 1.00 1.00 1.00 1.00 0.49 0.57 0.09 0.09 0.41 0.41 0.49 0.61
Controls
N, Newfoundland; P type I, psoriasis type I (early-onset); P type II, psoriasis type II (late-onset); PHWE, P value for Hardy–Weinberg equilibrium; PsA, psoriatic arthritis; PV, psoriasis vulgaris. *The disease types can vary based on different classifications; the alternative diagnosis is given in brackets.
psoriasis overall (OR = 2.06, 95% CI = 1.45–2.94, P < 0.001), and the heterogeneity was significant (v² = 43.24, P < 0.001, I2 = 65.3%). When stratified for ethnic origin, A carriers in white populations were found to have a slightly higher risk [OR = 2.40, 95% CI = 1.44–3.99, P value for odds ratio (POR) = 0.001; v² = 33.11, P-value for heterogeneity (Phet)< 0.001, I2 = 75.8%] than East Asian carriers (OR = 1.72, 95% CI = 1.16–2.56, POR < 0.01; v² = 5.27, Phet = 0.38, I2 = 5.1%). The forest plot is shown in Fig. 1a. Analyzing for age of onset, carriage of the A allele was found to be significantly associated with psoriasis type I (OR = 3.68, 95% CI = 2.17–6.24, POR < 0.001; v² = 17.67, Phet < 0.01, I2 = 66.0%) (Fig. 1b) but not with type II (OR = 0.98, 95% CI = 0.56–1.70, POR = 0.93; v² = 5.81, Phet = 0.21, I2 = 31.2%) (Fig. 1c). The association between the 238G/A polymorphism and psoriasis risk was similar for both PV (OR = 2.55, 95% CI = 1.40–4.67, POR = 0.002; v² = 37.90, Phet < 0.001, I2 = 78.9%) (Fig. 1d) and
838
Clinical and Experimental Dermatology (2013) 38, pp836–844
PsA (OR = 2.23, 95% CI = 1.69–2.95, POR < 0.001; v² = 7.74, Phet = 0.26, I2 = 22.5%) (Fig. 1e). Evaluation of heterogeneity in controls showed that the frequencies of the A and G alleles and the GA and GG genotypes could substantially influence the initial overall heterogeneity (P < 0.003, adjusted R2 = 80– 100%), and that the A and G alleles and the GG genotype partly contributed to the heterogeneity among cases (P < 0.029, adjusted R2 = 40–50%). For psoriasis type I, the sources of heterogeneity in controls were the frequencies of A and G alleles and the GA and GG genotypes (P < 0.013, adjusted R2 = 100%, respectively). For PV, the frequencies of the A and G allele in populations, the frequencies of the GG, GA and AA genotypes in controls, and the frequencies of GG genotype in cases were separate causes of heterogeneity (P < 0.03, adjusted R2 > 68%). Sensitivity analyses showed that there was no significant influence of any one study in meta-analysis estimation, and no evidence of publication bias was found in the overall study (P = 0.89 for Begg test and
ª 2013 British Association of Dermatologists
TNF-a polymorphism and psoriasis Y. Jia et al.
Table 2 Characteristics of the included studies on the tumour necrosis factor-a
308G/A (rs1800629) polymorphism. Genotypes n
Author
Year
Region
Ethnicity
Hohler6
1997
Germany
White
Reich12
1999
Germany
White
Hamamoto13 Craven14
2000 2001
Japan UK
East Asian White
Gonzalez15 AI-Heresh16 Li17
2001 2002 2002
Spain UK China
White White East Asian
Zhang18
2002
China
East Asian
Kim19 Tsunemi7 Long8
2003 2003 2004
Korea Japan China
East Asian East Asian East Asian
Mossner9 Baran20
2005 2006
Germany Poland
White White
Rahman10 Rahman10 Nedoszytko21
2006 2006 2007
Canada (N) Canada (Toronto) Poland
White White White
Reich3
2007
Germany
White
Settin22 Magalhaes23 Gallo11
2009 2010 2012
Egypt Brazil Spain
White White White
Classification of psoriasis P type PsA P type P type PsA P type P type PsA PsA P type P type P type P type PV PV P type P type PV P type P type PsA PsA P type P type PV PsA PV P type PV
I (PV)* I (PV) II (PV) I II
I II I (PV) II (PV)
I (PV) II (PV) I (PV) II (PV)
I (PV) II (PV)
I
Samples, n
Cases
Cases
Controls
GG
GA
AA
GG
GA
AA
PHWE
60 62 100 51 20 48 33 52 124 77 46 65 22 103 163 48 29 239 54 24 237 203 134 32 375 375 46 69 84
99 99 123 123 87 66 66 73 101 165 165 128 128 125 96 82 82 135 74 74 103 101 65 65 376 376 98 70 76
52 47 77 36 20 39 24 44 79 57 35 56 21 96 161 46 27 197 45 17 146 144 114 27 283 275 9 57 70
7 11 23 15 0 8 6 8 38 19 11 9 1 7 2 2 2 41 9 7 75 53 20 4 76 76 28 12 14
1 4 0 0 0 1 3 0 7 1 0 0 0 0 0 0 0 1 0 0 4 6 0 1 9 10 9 0 0
73 73 79 79 83 44 44 58 71 104 104 112 112 107 92 71 71 95 57 57 67 69 46 46 253 253 6 53 60
20 20 42 42 3 21 21 15 27 61 61 16 16 18 4 11 11 39 16 16 33 25 18 18 107 107 81 17 16
6 6 2 2 1 1 1 0 3 0 0 0 0 0 0 0 0 1 1 1 3 7 1 1 10 10 11 0 0
0.04 0.04 0.30 0.30 0.11 0.76 0.76 0.90 1.00 < 0.01 < 0.01 1.00 1.00 1.00 1.00 1.00 1.00 0.29 1.00 1.00 0.98 0.08 1.00 1.00 0.93 0.93 0.01 0.66 0.83
Controls
N, Newfoundland; P type I, psoriasis type I (early-onset); P type II, psoriasis type II (late-onset); PHWE, P value for Hardy–Weinberg equilibrium; PsA, psoriatic arthritis; PV, psoriasis vulgaris. *The disease types can vary based on different classifications; the alternative diagnosis is given in brackets.
P = 0.81 for Egger test). The Begg is funnel plot shown in Fig. 1f. Meta-analysis of
308G/A
The frequencies of the A allele ranged from 0.02 to 0.53 in controls. The pooled examination of the -308G/A polymorphism using the dominant model (AA/GA vs. GG, as the AA genotype was missing in some studies) showed that carriage of the A allele was associated with a reduced risk of psoriasis overall (OR = 0.68, 95% CI = 0.59–0.79, P < 0.001), and no observable heterogeneity between studies was found (v² = 19.71, P = 0.35, I2 = 8.7%). When stratified for ethnic origin, there was no significant difference between the white (OR = 0.71, 95% CI = 0.61–0.83, POR < 0.001; v² = 14.56, Phet = 0.22,
ª 2013 British Association of Dermatologists
I2 = 22.4%) and East Asian populations (OR = 0.55, 95% CI = 0.38–0.78, POR = 0.001; v² = 2.8, Phet = 0.74, I2 = 0.0%) (Fig. 2a). A significant association was found between carriage of the A allele and the incidence of psoriasis type I (OR = 0.56, 95% CI = 0.43–0.72, POR < 0.001; v² = 4.65, Phet = 0.79, I2 = 0.0%) (Fig. 2b). A significant association with carriage of the A allele was also found for psoriasis type II (OR = 0.66, 95% CI = 0.455 –0.94, POR = 0.02; v² = 3.52, Phet = 0.74, I2 = 0.0%) (Fig. 2c); however the removal of the studies by Li17 and Nedoszytko et al.21 changed the results (the pooled ORs were 0.70 with 95% CI = 0.46–1.06, and 0.69 with 95% CI = 0.47–1.01, respectively). There was a significant association between carriage of the A allele and the incidence of PV (OR = 0.57, 95% CI = 0.47–0.70, POR < 0.001; v² = 8.03, Phet =
Clinical and Experimental Dermatology (2013) 38, pp836–844
839
TNF-a polymorphism and psoriasis Y. Jia et al.
(a)
(b) %
Study OR (95% CI)
ID Caucasian Hohler (1997) Gonzalez (2001) AI−Heresh (2002) Mossner (2005) Rahman (2006) Rahman (2006) Nedoszytko (2007) Reich (2007) Gallo (2012) Subtotal (I−squared = 75.8%, P = 0.000) . Asian Hamamoto (2000) Li (2002) Zhang (2002) Kim (2003) Tsunemi (2003) Long (2004) Subtotal (I−squared = 5.1%, P = 0.384) . Latino Magalhaes (2010) Subtotal (I−squared = .%, P = .) . Overall (I−squared = 65.3%, P = 0.000)
Study
Weight
7.15 (3.05, 16.80) 3.63 (0.89, 14.77) 1.54 (0.74, 3.22) 3.21 (1.65, 6.23) 2.41 (1.17, 5.00) 1.75 (0.85, 3.60) 6.78 (2.02, 22.76) 2.85 (1.92, 4.22) 0.37 (0.17, 0.85) 2.40 (1.44, 3.99)
6.61 3.99 7.32 7.80 7.39 7.41 4.75 9.50 6.84 61.61
0.45 (0.02, 8.75) 1.42 (0.75, 2.67) 2.28 (1.07, 4.83) 1.25 (0.57, 2.75) 1.49 (0.28, 7.82) 5.35 (1.46, 19.59) 1.72 (1.16, 2.56)
1.26 8.00 7.23 6.99 3.19 4.39 31.05
1.17 (0.56, 2.45) 1.17 (0.56, 2.45)
7.34 7.34
2.06 (1.45, 2.94)
100.00
%
ID
OR (95% CI)
Caucasian Hohler (1997)
8.17 (3.23, 20.67) 13.45
Reich (2002)
4.01 (2.39, 6.75)
Nedoszytko (2007)
8.79 (2.61, 29.68) 10.37
Subtotal (I−squared = 24.3%, P = 0.267)
5.52 (3.23, 9.43)
1
18.79 42.62
. Asian Li (2002)
2.32 (1.19, 4.51)
16.81
Zhang (2002)
2.88 (1.32, 6.31)
15.26
Long (2004)
8.78 (2.33, 33.03) 9.41
Subtotal (I−squared = 36.1%, P = 0.209)
3.19 (1.71, 5.96)
41.48
Magalhaes (2010)
1.17 (0.56, 2.45)
15.90
Subtotal (I−squared = .%, P = .)
1.17 (0.56, 2.45)
15.90
3.68 (2.17, 6.24)
100.00
. Latino
. Overall (I−squared = 66.0%, P = 0.007)
0.0234
Weight
0.0303
1
33
(d)
42.7
Study ID
(c)
OR (95% CI)
% Weight
Caucasian
Study ID
OR (95% CI)
% Weight
Asian
Hohler (1997)
8.17 (3.23, 20.67) 11.20
Mossner (2005)
3.21 (1.65, 6.23)
Nedoszytko (2007)
6.78 (2.02, 22.76) 9.43
Reich (2007)
3.64 (2.39, 5.55)
14.19
Gallo (2012)
0.37 (0.17, 0.85)
11.92
Subtotal (I−squared = 87.6%, P = 0.000)
2.91 (1.14, 7.42)
59.62
Zhang (2002)
12.87
Li (2002)
0.28 (0.06, 1.24)
37.48
Zhang (2002)
0.81 (0.17, 3.86)
14.59
Long (2004)
0.94 (0.09, 9.42)
5.91
Subtotal (I−squared = 0.0%, P = 0.530)
0.48 (0.19, 1.26)
57.98
2.28 (1.07, 4.83)
12.32
Kim (2003)
1.25 (0.57, 2.75)
12.07
Reich (2002)
1.87 (0.89, 3.94)
34.53
Tsunemi (2003)
1.49 (0.28, 7.82)
7.07
Nedoszytko (2007)
0.67 (0.07, 6.68)
7.49
Long (2004)
5.35 (1.46, 19.59) 8.93
42.02
Subtotal (I−squared = 21.6%, P = 0.281)
2.02 (1.15, 3.57)
40.38
2.55 (1.40, 4.67)
100.00
. Asian
. Caucasian
Subtotal (I−squared = 0.0%, P = 0.402)
1.66 (0.82, 3.36)
.
. Overall (I−squared = 31.2%, P = 0.214)
0.98 (0.56, 1.70)
0.0636
1
Overall (I−squared = 78.9%, P = 0.000)
100.00
0.0439
15.7
4
% OR (95% CI)
Weight
Hohler (1997)
6.26 (2.46, 15.94)
5.22
Hamamoto (2000)
0.45 (0.02, 8.75)
2.42
Gonzalez (2001)
3.63 (0.89, 14.77)
3.09
AI−Heresh (2002)
1.54 (0.74, 3.22)
16.68
Rahman (2006)
2.41 (1.17, 5.00)
15.63
Rahman (2006)
1.75 (0.85, 3.60)
17.19
Reich (2007)
2.13 (1.37, 3.32)
39.79
Overall (I−squared = 22.5%, P = 0.258)
2.23 (1.69, 2.95)
100.00
2
logOR
ID
22.8
Begg’s funnel plot with pseudo 95% confidence limits
(f)
(e) Study
1
0
–2 0.0234
1
42.7
0
0.5
1
1.5
s.e. of: logOR
Figure 1 Meta-analysis of the tumour necrosis factor (TNF)-a 238G/A polymorphism (rs361525) with the dominant model (AA/GA vs. GG). (a–e) Forest plots for (a) psoriasis overall with stratification for ethnic origin; (b) psoriasis type I (early-onset); (c) psoriasis type II (late-onset); (d) psoriasis vulgaris (PV); and (e) psoriatic arthritis (PsA). (f) Begg funnel plot for psoriasis overall.
0.63, I2 = 0.0%) (Fig. 2d). However, there was no association found with PsA (OR = 0.85, 95% CI = 0.69–1.04, POR = 0.12; v² = 5.30, Phet = 0.51,
840
Clinical and Experimental Dermatology (2013) 38, pp836–844
I2 = 0.0%) (Fig. 2e), and omission of the study by AIHeresh et al.16 changed the results (the pooled OR was 0.79 with 95% CI = 0.63–0.98).
ª 2013 British Association of Dermatologists
TNF-a polymorphism and psoriasis Y. Jia et al.
(b)
(a) Study ID
OR (95% CI)
Caucasian Hohler (1997) Craven (2001) Gonzalez (2001) AI−Heresh (2002) Mossner (2005) Baran (2006) Rahman (2006) Rahman (2006) Nedoszytko (2007) Reich (2007) Settin (2009) Magalhaes (2010) Gallo (2012) Subtotal (I−squared = 22.4%, P = 0.217) . Asian Hamamoto (2000) Li (2002) Zhang (2002) Kim (2003) Tsunemi (2003) Long (2004) Subtotal (I−squared = 0.0%, P = 0.737) . Overall (I−squared = 8.7%, P = 0.349)
0.65 (0.34, 1.23) 0.57 (0.27, 1.19) 0.70 (0.27, 1.81) 1.35 (0.77, 2.37) 0.51 (0.31, 0.83) 0.87 (0.40, 1.87) 1.01 (0.62, 1.64) 0.88 (0.53, 1.48) 0.43 (0.22, 0.85) 0.66 (0.50, 0.88) 0.27 (0.09, 0.81) 0.66 (0.29, 1.50) 0.75 (0.34, 1.66) 0.71 (0.61, 0.83)
0.0234
1
% Weight
Study ID
0.37 8.55 2.51 3.32 1.09 2.22 18.06
0.68 (0.59, 0.79)
100.00
42.7
% Weight
Caucasian
5.11 4.14 2.32 4.62 9.24 3.04 7.03 6.65 5.09 26.06 2.52 3.06 3.07 81.94
0.45 (0.02, 8.75) 0.57 (0.34, 0.96) 0.91 (0.39, 2.11) 0.43 (0.17, 1.08) 0.29 (0.05, 1.59) 0.35 (0.11, 1.16) 0.55 (0.38, 0.79)
OR (95% CI)
Hohler (1997)
0.43 (0.18, 1.03)
10.91
Reich (1999)
0.54 (0.30, 0.97)
19.49
Craven (2001)
0.46 (0.19, 1.12)
Baran (2006)
0.67 (0.27, 1.65)
7.67
Nedoszytko (2007)
0.42 (0.21, 0.87)
13.96
Magalhaes (2010)
0.66 (0.29, 1.50)
8.94
Subtotal (I−squared = 0.0%, P = 0.947)
0.52 (0.38, 0.71)
70.62
Li (2002)
9.65
. Asian 0.60 (0.33, 1.09)
18.43
Zhang (2002)
1.13 (0.47, 2.71)
5.96
Long (2004)
0.28 (0.06, 1.32)
4.99
Subtotal (I−squared = 25.9%, P = 0.259)
0.65 (0.41, 1.04)
29.38
0.56 (0.43, 0.72)
100.00
. Overall (I−squared = 0.0%, P = 0.794)
0.0595
1
16.8
(d) Study
(c)
%
ID
Study ID
OR (95% CI)
% Weight
OR (95% CI)
Weight
Caucasian
Caucasian
Hohler (1997)
0.43 (0.18, 1.03)
6.72
Mossner (2005)
0.51 (0.31, 0.83)
16.65
Baran (2006)
0.87 (0.40, 1.87)
5.48
Reich (1999)
0.75 (0.37, 1.52)
24.14
Nedoszytko (2007)
0.43 (0.22, 0.85)
9.16
Craven (2001)
0.75 (0.30, 1.88)
14.14
Reich (2007)
0.65 (0.47, 0.90)
35.45
Baran (2006)
1.38 (0.49, 3.88)
7.82
Settin (2009)
0.27 (0.09, 0.81)
Nedoszytko (2007)
0.45 (0.15, 1.34)
14.03
Gallo (2012)
0.75 (0.34, 1.66)
5.53
60.13
Subtotal (I−squared = 0.0%, P = 0.512)
0.58 (0.47, 0.72)
83.53
Zhang (2002)
0.91 (0.39, 2.11)
4.53
Kim (2003)
0.43 (0.17, 1.08)
5.99
Tsunemi (2003)
0.29 (0.05, 1.59)
1.96
Subtotal (I−squared = 0.0%, P = 0.536)
0.76 (0.49, 1.19)
4.54
.
.
Asian
Asian Zhang (2002)
0.33 (0.04, 2.65)
5.94
Li (2002)
0.54 (0.25, 1.13)
26.83
Long (2004)
0.48 (0.10, 2.30)
7.10
Long (2004)
0.35 (0.11, 1.16)
3.99
39.87
Subtotal (I−squared = 0.0%, P = 0.439)
0.53 (0.31, 0.88)
16.47
0.57 (0.47, 0.70)
100.00
Subtotal (I−squared = 0.0%, P = 0.912)
0.50 (0.26, 0.94)
.
. Overall (I−squared = 0.0%, P = 0.741)
0.66 (0.46, 0.94)
0.0419
1
Overall (I−squared = 0.0%, P = 0.626)
100.00
0.0513
23.9
4
% OR (95% CI)
Weight
Hohler (1997)
0.90 (0.43, 1.87)
7.63
Hamamoto (2000)
0.45 (0.02, 8.75)
0.85
Gonzalez (2001)
0.70 (0.27, 1.81)
5.31
AI−Heresh (2002)
1.35 (0.77, 2.37)
10.59
Rahman (2006)
1.01 (0.62, 1.64)
16.11
Rahman (2006)
0.88 (0.53, 1.48)
15.24
Reich (2007)
0.68 (0.49, 0.94)
44.26
Overall (I−squared = 0.0%, P = 0.506)
0.85 (0.69, 1.04)
100.00
2
logOR
ID
19.5
Begg’s funnel plot with pseudo 95% confidence limits
(f)
(e) Study
1
0
–2
–4 0.0234
1
42.7
0
0.5
1
1.5
s.e. of: logOR
Figure 2 Meta-analysis of the tumour necrosis factor (TNF)-a 308G/A polymorphism (rs1800629) with the dominant model (AA/GA vs. GG). (a–e) Forest plots for (a) psoriasis overall with stratification for ethnic origin; (b) psoriasis type I (early-onset); (c) psoriasis type II (late-onset); (d) psoriasis vulgaris (PV); and (e) psoriatic arthritis (PsA). (f) Begg funnel plot for psoriasis overall.
Sensitivity analyses found no significant influence of any one study to the incidence of psoriasis overall, psoriasis type I or PV. No evidence of publication bias
ª 2013 British Association of Dermatologists
was found for the overall study (P = 0.23 for Begg test and P = 0.22 for Egger test). The Begg funnel plot is shown in Fig. 2f.
Clinical and Experimental Dermatology (2013) 38, pp836–844
841
TNF-a polymorphism and psoriasis Y. Jia et al.
Discussion The present meta-analysis identified that carriers of the A allele (AA/GA) of the TNF-a 238G/A polymorphism had an increased risk of psoriasis overall, and that there were no significant differences between white and East Asian populations or between PV and PsA. However, when we divided psoriasis into two groups based on age of onset, we found that this increased risk was present only in psoriasis type I but not in type II, which was not consistent with the results of Li et al.4 It suggests that the susceptibility difference of the TNF-a 238G/A variant may be associated with age of onset only and not with classifications of either clinical features or ethnic origin. In addition, the positive associations of the A allele with psoriasis overall, as well as with PV and PsA, are possibly due to the large numbers of patients with type I psoriasis, who made up almost three-quarters of all patients. There was obvious heterogeneity between studies for psoriasis overall, psoriasis type I and PV, which was mainly related to the frequencies of the A and G alleles and the GA and GG genotypes in populations. Sensitivity analyses indicated that our results for the 238G/A polymorphism were statistically reliable. The meta-analysis of the 308G/A polymorphism showed that carriage of the A allele conferred a decreased risk for psoriasis in the overall study and each group except for PsA, and no heterogeneity between studies was found. However, the results for the type II psoriasis and PsA groups were not statistically reliable because the pooled OR had a significant difference, due to some study omissions. The morbidity rate of PV is much higher than that of PsA; however, in our meta-analysis, there was no significant difference between PV and PsA for the association of the TNF-a 238G/A polymorphism with psoriasis risk. Thus, the differences in morbidity rate between these two conditions may be attributable to other factors. For example, the TNF-857C/T polymorphism has been considered to be a genetic factor independent of the PSORS1 risk gene, which can influence the development of PsA.3 How the different roles of TNF-a –238 and –308 variants affect psoriasis risk remains unknown. The risk most probably relates to the variant-mediated abnormal gene expression;4 however, the functional effects of the two variants are still inconsistent.2,5,24 The human leucocyte antigen (HLA) region exhibits
842
Clinical and Experimental Dermatology (2013) 38, pp836–844
extensive linkage disequilibrium across the whole region, and TNF-a is known to be in strong linkage disequilibrium with some HLA haplotypes.6,15 Hence, it is difficult to tell whether the relationship between the polymorphisms and psoriasis is caused by the direct influence of the polymorphisms themselves or by linkage disequilibrium with other HLA genes. Inevitably, the data included in the present metaanalysis had some limitations. First, some controls were not in HWE, although this did not contribute to the heterogeneity seen in this meta-analysis. Second, some studies were not included in this meta-analysis because of the full text was unavailable or there were insufficient data, and some of the included studies were not adequate, especially in the subgroups. This probably led to an inaccurate estimate of publication bias, even though the relative statistical tests did not show this because the sensitivities of both the Begg and Egger tests were generally low in this meta-analysis, based on < 20 trials.25
Conclusion This meta-analysis supported an overall increased psoriasis risk associated with the TNF-a 238G/A polymorphism and a reduced risk with the 308G/ A polymorphism. The 238G/A polymorphism only had an effect in early-onset patients, and there were no significant differences in ethnicity or classifications of clinical features. For 308G/A, there were also no significant differences between white and East Asian populations; however, the accuracy and reliability of the results, especially in psoriasis type II and PsA, should be confirmed by well-designed studies with a larger sample size and disease classifications. Further, relative functional studies also should be performed to better explore the complex and comprehensive roles of the TNF-a promoter polymorphism.
What’s already known about this topic?
•
TNF-a may be involved in the aetiology of psoriasis. • The association between TNF-a -238G/A and 308G/A polymorphisms and psoriasis risk is still ambiguous.
ª 2013 British Association of Dermatologists
TNF-a polymorphism and psoriasis Y. Jia et al.
What does this study add?
•
There is an an overall increased psoriasis risk was associated with the TNF-a 238G/A polymorphism. • There is a reduced psoriasis risk with the TNFa 308G/A polymorphism.
References 1 Elder JT, Bruce AT, Gudjonsson JE et al. Molecular dissection of psoriasis: integrating genetics and biology. J Invest Dermatol 2010; 130: 1213–1226. 2 Elahi MM, Asotra K, Matata BM, Mastana SS. Tumor necrosis factor alpha -308 gene locus promoter polymorphism. An analysis of association with health and disease. Biochim Biophys Acta 2009; 1792: 163–172. 3 Reich K, Huffmeier U, Konig IR et al. TNF polymorphisms in psoriasis. Association of psoriatic arthritis with the promoter polymorphism TNF*-857 independent of the PSORS1 risk allele. Arthritis Rheum 2007; 56: 2056–2064. 4 Li C, Wang G, Gao Y et al. TNF-alpha gene promoter 238G>A and 308G>A polymorphisms alter risk of psoriasis vulgaris: a meta-analysis. J Invest Dermatol 2007; 127: 1886–1892. 5 Reich K, Mossner R, Konig IR et al. Promoter polymorphisms of the genes encoding tumor necrosis factor-alpha and interleukin-1beta are associated with different subtypes of psoriasis characterized by early and late disease onset. J Invest Dermatol 2002; 118: 155–163. 6 Hohler T, Kruger A, Schneider PM et al. A TNF-alpha promoter polymorphism is associated with juvenile onset psoriasis and psoriatic arthritis. J Invest Dermatol 1997; 109: 562–565. 7 Tsunemi Y, Nishibu A, Saeki H et al. Lack of association between the promoter polymorphisms at positions 308 and 238 of the tumor necrosis factor alpha gene and psoriasis vulgaris in Japanese patients. Dermatology 2003; 207: 371–374. 8 Long F, Sun C, Deng D et al. TNF-238A is associated with juvenile onset psoriasis in patients of Han population in southwest China. J Dermatol Sci 2004; 36: 109–111. 9 Mossner R, Kingo K, Kleensang A et al. Association of TNF 238 and 308 promoter polymorphisms with psoriasis vulgaris and psoriatic arthritis but not with pustulosis palmoplantaris. J Invest Dermatol 2005; 124: 282–284. 10 Rahman P, Siannis F, Butt C et al. TNF-alpha polymorphisms and risk of psoriatic arthritis. Ann Rheum Dis 2006; 65: 919–923.
ª 2013 British Association of Dermatologists
11 Gallo E, Cabaleiro T, Rom an M et al. Study of genetic polymorphisms in the tumor necrosis factor a promoter region in Spanish patients with psoriasis. Actas Dermosifiliogr 2012; 103: 301–307. 12 Reich K, Westphal G, Schulz T et al. Combined analysis of polymorphisms of the tumor necrosis factor-alpha and interleukin-10 promoter regions and polymorphic xenobiotic metabolizing enzymes in psoriasis. J Invest Dermatol 1999; 113: 214–220. 13 Hamamoto Y, Tateno H, Ishida T, Muto M. Lack of association between promoter polymorphism of the tumor necrosis factor-alpha gene and psoriatic arthritis in Japanese patients. J Invest Dermatol 2000; 115: 1162–1164. 14 Craven NM, Jackson CW, Kirby B et al. Cytokine gene polymorphisms in psoriasis. Br J Dermatol 2001; 144: 849–853. 15 Gonzalez S, Brautbar C, Martinez-Borra J et al. Polymorphism in MICA rather than HLA-B/C genes is associated with psoriatic arthritis in the Jewish population. Hum Immunol 2001; 62: 632–638. 16 Al-Heresh AM, Proctor J, Jones SM et al. Tumour necrosis factor-alpha polymorphism and the HLACw*0602 allele in psoriatic arthritis. Rheumatology (Oxford) 2002; 41: 525–530. 17 Li B. Tumor Necrosis Factor-Α Gene Polymorphism and Psoriasis. PhD Thesis. Shenyang: China Medical University, 2002. 18 Zhang AP. Experimental Studies on HLA Genotyping in Patients with Psoriasis Vulgaris in Chinese Han Population. PhD Thesis. Nanjing: Nanjing Medical University, 2002. 19 Kim TG, Pyo CW, Hur SS et al. Polymorphisms of tumor necrosis factor (TNF) alpha and beta genes in Korean patients with psoriasis. Arch Dermatol Res 2003; 295: 8–13. 20 Baran W, Szepietowski JC, Mazur G, Baran E. A 308 promoter polymorphism of tumor necrosis factor alpha gene does not associate with the susceptibility to psoriasis vulgaris. No difference either between psoriasis type I and type II patients. Acta Dermatovenerol Alp Panonica Adriat 2006; 15: 113–118. 21 Nedoszytko B, Szczerkowska-Dobosz A, Zablotna M et al. Associations of promoter region polymorphisms in the tumour necrosis factor-alpha gene and early-onset psoriasis vulgaris in a northern Polish population. Br J Dermatol 2007; 157: 165–167. 22 Settin A, Hassan H, El-Baz R, Hassan T. Association of cytokine gene polymorphisms with psoriasis in cases from the Nile Delta of Egypt. Acta Dermatovenerol Alp Panonica Adriat 2009; 18: 105–112. 23 Magalhaes RF, Biral AC, Pancoto JA et al. Human leukocyte antigen (HLA) and single nucleotide polymorphisms (SNPs) tumor necrosis factor (TNF)-alpha -238 and -308 as genetic markers of susceptibility to
Clinical and Experimental Dermatology (2013) 38, pp836–844
843
TNF-a polymorphism and psoriasis Y. Jia et al.
psoriasis and severity of the disease in a long-term follow-up Brazilian study. Int J Dermatol 2010; 49: 1133–1140. 24 Kaluza W, Reuss E, Grossmann S et al. Different transcriptional activity and in vitro TNF-alpha production in psoriasis patients carrying the TNF-alpha
844
Clinical and Experimental Dermatology (2013) 38, pp836–844
238A promoter polymorphism. J Invest Dermatol 2000; 114: 1180–1183. 25 Sterne JA, Gavaghan D, Egger M. Publication and related bias in meta-analysis: power of statistical tests and prevalence in the literature. J Clin Epidemiol 2000; 53: 1119–1129.
ª 2013 British Association of Dermatologists
Clinical dermatology • Original article
Clinical and Experimental Dermatology
Association of the tumour necrosis factor-a polymorphisms rs361525 and rs1800629 with susceptibility to psoriasis: a meta-analysis Y. Jia,1 H. J. Qin,2 J. X. Zhang,1 X. L. Liu1 and L. J. Li1 1
First Affiliated Hospital and 2School of Forensic Medicine, Henan University of Science and Technology, Luoyang, China
doi:10.1111/ced.12136
Summary
Background. Evidence has suggested that tumour necrosis factor (TNF)-a may be involved in the aetiology of psoriasis, but the underlying association of the TNF-a polymorphisms 238G/A (rs361525) and 308G/A (rs1800629) with the risk of psoriasis is still unconfirmed. Aim. This meta-analysis was performed to determine whether the TNF-a 238G/A and 308G/A polymorphisms are associated with susceptibility to psoriasis. Methods. Eligible studies were identified by searching PubMed, EMBASE, CNKI (China National Knowledge Infrastructure), CBM (Chinese biomedical literature database) and WANFANG databases within a range of published years from 1990 to August 2012. The odds ratio (OR) and 95% confidence interval (CI) were used to assess the different associations. Results. In total, 17 studies with 2847 cases and 2222 controls were found for 238G/A and 20 studies with 2975 cases and 2243 controls for 308G/A. The pooled results showed an overall increased risk of psoriasis for the 238G/A polymorphism (OR = 2.06, 95% CI = 1.45–2.94, P < 0.001 for AA/GA vs. GG) and a reduced psoriasis risk with the 308G/A polymorphism (OR = 0.68, 95% CI = 0.59–0.79, P < 0.001 for AA/GA vs. GG). This association was only present in early-onset psoriasis (OR = 3.68, 95% CI = 2.17–6.24, P < 0.001 for 238G/A; OR = 0.56, 95% CI = 0.43–0.72, P < 0.001 for 308G/A), whereas there was no association (OR = 0.98, 95% CI = 0.56–1.70, P = 0.92 for 238G/A) or a unreliable association (OR = 0.66, 95% CI = 0.46–0.94, P = 0.02 for 308G/A) in late-onset psoriasis. Conclusions. This meta-analysis suggests that the TNF-a –238 and –308 promoter polymorphisms may play different roles in conferring susceptibility to psoriasis. Functional and well-designed studies should be conducted to confirm these results.
Introduction Psoriasis is a common, chronic inflammatory skin disease, affecting 2–3% of the global population. Two Correspondence: Dr Lijun Li, First Affiliated Hospital, Henan University of Science and Technology, Luoyang 471003, China E-mail: [email protected] Conflict of interest: none declared. The first two authors contributed equally to this work and should be considered joint first authors. Accepted for publication 16 December 2012
836
Clinical and Experimental Dermatology (2013) 38, pp836–844
types are recognised, based on age of onset. Type I (early-onset) psoriasis has an age of onset of < 40 years, and tends to be familial and severe. Type II (late-onset) occurs after the age of 40 years, and is more likely to be sporadic.1 Psoriasis is also divided based on clinical symptoms into psoriasis vulgaris (PV), pustular psoriasis, erythrodermic psoriasis, and psoriatic arthritis (PsA). Approximately 5–30% of patients with psoriasis may also develop PsA. Psoriasis is a T-cell-mediated disease, driven at least partly by a positive feedback loop from activated T cells to antigenpresenting cells, which is mediated by interferon-c,
ª 2013 British Association of Dermatologists
TNF-a polymorphism and psoriasis Y. Jia et al.
interleukin-1 and tumour necrosis factor (TNF)-a.1 TNF-a is a proinflammatory cytokine that has been implicated in the pathogenesis of a large number of human diseases.2 TNF-a levels are increased in psoriatic lesions, and biological drugs that block TNF-a are highly effective treatments for psoriasis, strongly supporting the role for TNF-a in the disease pathogenesis.1 Genome-wide association studies have identified several psoriasis susceptibility loci, including a locus on the short arm of chromosome at 6p21 (PSORS1).3 The TNF-a gene is also located nearby (6p21.3).2,4 Several polymorphisms have been identified in the TNF-a promoter region. The single nucleotide polymorphisms 238G/A (rs361525) and 308G/A (rs1800629) have been shown to directly affect TNF-a expression, and are therefore candidates for genetic psoriasis research.2,5 Several studies have explored the potential association between TNF-a promoter polymorphisms and susceptibility to psoriasis, but the results were inconsistent.3,6–11 A meta-analysis was performed in 2007, but it focused only on PV and had a relatively small sample size.4 Hence, we conducted a metaanalysis to investigate the proposed correlation more thoroughly.
Methods Literature search strategy
Electronic searches were conducted of PubMed, EMBASE, CNKI (China National Knowledge Infrastructure), CBM (Chinese biomedical literature database) and WANFANG databases (updated to August 6, 2012), using the terms ‘psoriasis’, ‘psoriatic arthritis’, ‘tumour necrosis factoralpha’ or ‘TNF-alpha’ and ‘polymorphism’. No restrictions were placed on language. Only published journal articles and academic dissertations were included. The reference lists of the research papers were also searched to identify other relevant publications. Inclusion and exclusion criteria
All studies included in the meta-analysis were required to meet the following criteria: (i) case–control study; (ii) evaluating the association of TNF-a rs361525 (-238G/A) and/or rs1800629 (-308G/A) polymorphisms with psoriasis risk; and (iii) presenting sufficient available data to assess odds ratio (OR) and 95% CI. Studies with insufficient or duplicate data were excluded. For overlapping and republished studies, the first study published or the one with the largest sample size (depending on the data) were included.
ª 2013 British Association of Dermatologists
Data extraction
All data were extracted independently by two reviewers (JY and QH) according to the prespecified selection criteria, and were reviewed and checked by a third investigator (ZJ). Disagreements about inclusion of studies and interpretation of data were resolved by discussion. Statistical analysis
Hardy–Weinberg equilibrium (HWE) was assessed in control groups using the Fisher exact test, with R software (version 2.15.0; R Foundation for Statistical Computing, Vienna, Austria) before statistical analysis. Dichotomous data are presented as OR with 95% CI. Statistical heterogeneity was measured using the Qstatistic and I2 test. A fixed-effects model was used to estimate the summary OR when there was no heterogeneity; otherwise, the random effects model was used. To explore sources of heterogeneity across studies, logistic meta-regression analyses were conducted. Sensitivity analyses were performed to investigate the influence of individual studies on the summary-effect estimates. The Begg rank correlation method and Egger weighted regression method were also used to assess the publication bias. The statistical analyses were performed using Stata software (version 12.0; StataCorp, College Station, TX, USA), and P < 0.05 was considered significant.
Results Study characteristics
The search strategy retrieved 69 potentially relevant articles. Based on the inclusion criteria, only 20 articles,3,5–23 consisting of 21 studies (1 article was considered to have 2 studies with 2 different populations) were included in the meta-analysis. Of these 21 studies, 17 studied TNF-a 238G/A in 2847 cases and 2222 controls, and 20 studies assessed TNF-a 308G/A in 2975 cases and 2243 controls. The corresponding characteristics are shown in Tables 1 and 2. Meta-analysis of
238G/A
The frequencies of the A allele of 238G/A ranged from 0.01 to 0.17 in control subjects. Because the AA genotype was not included in some studies, the pooled examination was carried out on the dominant genetic model (AA/GA vs. GG). This showed that carriage of the A allele was associated with an increased risk of
Clinical and Experimental Dermatology (2013) 38, pp836–844
837
TNF-a polymorphism and psoriasis Y. Jia et al.
Table 1 Characteristics of the included studies on the tumour necrosis factor-a
238G/A (rs361525) polymorphism. Genotypes n
Author
Year
Region
Ethnicity
Hohler6
1997
Germany
White
Hamamoto13 Gonzalez15 AI-Heresh16 Li17
2000 2001 2002 2002
Japan Spain UK China
East Asian White White East Asian
Reich5
2002
Germany
White
Zhang18
2002
China
East Asian
Kim19 Tsunemi7 Long8
2003 2003 2004
Korea Japan China
East Asian East Asian East Asian
Mossner9 Rahman10 Rahman10 Nedoszytko21
2005 2006 2006 2007
Germany Canada (N) Canada (Toronto) Poland
White White White White
Reich3
2007
Germany
White
Magalhaes23 Gallo11
2010 2012
Brazil Spain
White White
Classification of psoriasis P type PsA PsA PsA PsA P type P type P type P type P type P type PV PV P type P type PV PsA PsA P type P type PV PsA P type PV
I (PV)*
I II I (PV) II (PV) I (PV) II (PV)
I (PV) II (PV)
I (PV) II (PV)
I
Samples, n
Cases
Cases
Controls
GG
GA
AA
GG
GA
AA
PHWE
60 62 20 52 124 77 46 156 75 65 22 103 163 48 29 239 237 203 134 32 375 375 69 81
99 99 87 73 101 165 165 345 345 128 128 125 96 82 82 135 103 101 65 65 376 376 70 71
37 42 20 45 101 56 44 114 64 48 20 89 158 36 28 182 181 164 94 31 273 303 48 70
18 20 0 7 23 19 1 41 11 17 2 14 5 9 1 55 45 35 35 1 97 60 21 10
5 0 0 0 0 2 1 1 0 0 0 0 0 3 0 2 2 0 5 0 3 5 0 1
92 92 83 70 88 142 142 316 316 114 114 111 94 79 79 123 93 90 62 62 338 338 51 50
7 7 4 3 12 20 20 29 29 12 12 13 2 3 3 12 9 10 2 2 32 32 19 18
0 0 0 0 1 3 3 0 0 2 2 1 0 0 0 0 1 1 1 1 2 2 0 3
1.00 1.00 1.00 1.00 0.77 0.12 0.12 1.00 1.00 0.14 0.14 0.72 1.00 1.00 1.00 1.00 0.49 0.57 0.09 0.09 0.41 0.41 0.49 0.61
Controls
N, Newfoundland; P type I, psoriasis type I (early-onset); P type II, psoriasis type II (late-onset); PHWE, P value for Hardy–Weinberg equilibrium; PsA, psoriatic arthritis; PV, psoriasis vulgaris. *The disease types can vary based on different classifications; the alternative diagnosis is given in brackets.
psoriasis overall (OR = 2.06, 95% CI = 1.45–2.94, P < 0.001), and the heterogeneity was significant (v² = 43.24, P < 0.001, I2 = 65.3%). When stratified for ethnic origin, A carriers in white populations were found to have a slightly higher risk [OR = 2.40, 95% CI = 1.44–3.99, P value for odds ratio (POR) = 0.001; v² = 33.11, P-value for heterogeneity (Phet)< 0.001, I2 = 75.8%] than East Asian carriers (OR = 1.72, 95% CI = 1.16–2.56, POR < 0.01; v² = 5.27, Phet = 0.38, I2 = 5.1%). The forest plot is shown in Fig. 1a. Analyzing for age of onset, carriage of the A allele was found to be significantly associated with psoriasis type I (OR = 3.68, 95% CI = 2.17–6.24, POR < 0.001; v² = 17.67, Phet < 0.01, I2 = 66.0%) (Fig. 1b) but not with type II (OR = 0.98, 95% CI = 0.56–1.70, POR = 0.93; v² = 5.81, Phet = 0.21, I2 = 31.2%) (Fig. 1c). The association between the 238G/A polymorphism and psoriasis risk was similar for both PV (OR = 2.55, 95% CI = 1.40–4.67, POR = 0.002; v² = 37.90, Phet < 0.001, I2 = 78.9%) (Fig. 1d) and
838
Clinical and Experimental Dermatology (2013) 38, pp836–844
PsA (OR = 2.23, 95% CI = 1.69–2.95, POR < 0.001; v² = 7.74, Phet = 0.26, I2 = 22.5%) (Fig. 1e). Evaluation of heterogeneity in controls showed that the frequencies of the A and G alleles and the GA and GG genotypes could substantially influence the initial overall heterogeneity (P < 0.003, adjusted R2 = 80– 100%), and that the A and G alleles and the GG genotype partly contributed to the heterogeneity among cases (P < 0.029, adjusted R2 = 40–50%). For psoriasis type I, the sources of heterogeneity in controls were the frequencies of A and G alleles and the GA and GG genotypes (P < 0.013, adjusted R2 = 100%, respectively). For PV, the frequencies of the A and G allele in populations, the frequencies of the GG, GA and AA genotypes in controls, and the frequencies of GG genotype in cases were separate causes of heterogeneity (P < 0.03, adjusted R2 > 68%). Sensitivity analyses showed that there was no significant influence of any one study in meta-analysis estimation, and no evidence of publication bias was found in the overall study (P = 0.89 for Begg test and
ª 2013 British Association of Dermatologists
TNF-a polymorphism and psoriasis Y. Jia et al.
Table 2 Characteristics of the included studies on the tumour necrosis factor-a
308G/A (rs1800629) polymorphism. Genotypes n
Author
Year
Region
Ethnicity
Hohler6
1997
Germany
White
Reich12
1999
Germany
White
Hamamoto13 Craven14
2000 2001
Japan UK
East Asian White
Gonzalez15 AI-Heresh16 Li17
2001 2002 2002
Spain UK China
White White East Asian
Zhang18
2002
China
East Asian
Kim19 Tsunemi7 Long8
2003 2003 2004
Korea Japan China
East Asian East Asian East Asian
Mossner9 Baran20
2005 2006
Germany Poland
White White
Rahman10 Rahman10 Nedoszytko21
2006 2006 2007
Canada (N) Canada (Toronto) Poland
White White White
Reich3
2007
Germany
White
Settin22 Magalhaes23 Gallo11
2009 2010 2012
Egypt Brazil Spain
White White White
Classification of psoriasis P type PsA P type P type PsA P type P type PsA PsA P type P type P type P type PV PV P type P type PV P type P type PsA PsA P type P type PV PsA PV P type PV
I (PV)* I (PV) II (PV) I II
I II I (PV) II (PV)
I (PV) II (PV) I (PV) II (PV)
I (PV) II (PV)
I
Samples, n
Cases
Cases
Controls
GG
GA
AA
GG
GA
AA
PHWE
60 62 100 51 20 48 33 52 124 77 46 65 22 103 163 48 29 239 54 24 237 203 134 32 375 375 46 69 84
99 99 123 123 87 66 66 73 101 165 165 128 128 125 96 82 82 135 74 74 103 101 65 65 376 376 98 70 76
52 47 77 36 20 39 24 44 79 57 35 56 21 96 161 46 27 197 45 17 146 144 114 27 283 275 9 57 70
7 11 23 15 0 8 6 8 38 19 11 9 1 7 2 2 2 41 9 7 75 53 20 4 76 76 28 12 14
1 4 0 0 0 1 3 0 7 1 0 0 0 0 0 0 0 1 0 0 4 6 0 1 9 10 9 0 0
73 73 79 79 83 44 44 58 71 104 104 112 112 107 92 71 71 95 57 57 67 69 46 46 253 253 6 53 60
20 20 42 42 3 21 21 15 27 61 61 16 16 18 4 11 11 39 16 16 33 25 18 18 107 107 81 17 16
6 6 2 2 1 1 1 0 3 0 0 0 0 0 0 0 0 1 1 1 3 7 1 1 10 10 11 0 0
0.04 0.04 0.30 0.30 0.11 0.76 0.76 0.90 1.00 < 0.01 < 0.01 1.00 1.00 1.00 1.00 1.00 1.00 0.29 1.00 1.00 0.98 0.08 1.00 1.00 0.93 0.93 0.01 0.66 0.83
Controls
N, Newfoundland; P type I, psoriasis type I (early-onset); P type II, psoriasis type II (late-onset); PHWE, P value for Hardy–Weinberg equilibrium; PsA, psoriatic arthritis; PV, psoriasis vulgaris. *The disease types can vary based on different classifications; the alternative diagnosis is given in brackets.
P = 0.81 for Egger test). The Begg is funnel plot shown in Fig. 1f. Meta-analysis of
308G/A
The frequencies of the A allele ranged from 0.02 to 0.53 in controls. The pooled examination of the -308G/A polymorphism using the dominant model (AA/GA vs. GG, as the AA genotype was missing in some studies) showed that carriage of the A allele was associated with a reduced risk of psoriasis overall (OR = 0.68, 95% CI = 0.59–0.79, P < 0.001), and no observable heterogeneity between studies was found (v² = 19.71, P = 0.35, I2 = 8.7%). When stratified for ethnic origin, there was no significant difference between the white (OR = 0.71, 95% CI = 0.61–0.83, POR < 0.001; v² = 14.56, Phet = 0.22,
ª 2013 British Association of Dermatologists
I2 = 22.4%) and East Asian populations (OR = 0.55, 95% CI = 0.38–0.78, POR = 0.001; v² = 2.8, Phet = 0.74, I2 = 0.0%) (Fig. 2a). A significant association was found between carriage of the A allele and the incidence of psoriasis type I (OR = 0.56, 95% CI = 0.43–0.72, POR < 0.001; v² = 4.65, Phet = 0.79, I2 = 0.0%) (Fig. 2b). A significant association with carriage of the A allele was also found for psoriasis type II (OR = 0.66, 95% CI = 0.455 –0.94, POR = 0.02; v² = 3.52, Phet = 0.74, I2 = 0.0%) (Fig. 2c); however the removal of the studies by Li17 and Nedoszytko et al.21 changed the results (the pooled ORs were 0.70 with 95% CI = 0.46–1.06, and 0.69 with 95% CI = 0.47–1.01, respectively). There was a significant association between carriage of the A allele and the incidence of PV (OR = 0.57, 95% CI = 0.47–0.70, POR < 0.001; v² = 8.03, Phet =
Clinical and Experimental Dermatology (2013) 38, pp836–844
839
TNF-a polymorphism and psoriasis Y. Jia et al.
(a)
(b) %
Study OR (95% CI)
ID Caucasian Hohler (1997) Gonzalez (2001) AI−Heresh (2002) Mossner (2005) Rahman (2006) Rahman (2006) Nedoszytko (2007) Reich (2007) Gallo (2012) Subtotal (I−squared = 75.8%, P = 0.000) . Asian Hamamoto (2000) Li (2002) Zhang (2002) Kim (2003) Tsunemi (2003) Long (2004) Subtotal (I−squared = 5.1%, P = 0.384) . Latino Magalhaes (2010) Subtotal (I−squared = .%, P = .) . Overall (I−squared = 65.3%, P = 0.000)
Study
Weight
7.15 (3.05, 16.80) 3.63 (0.89, 14.77) 1.54 (0.74, 3.22) 3.21 (1.65, 6.23) 2.41 (1.17, 5.00) 1.75 (0.85, 3.60) 6.78 (2.02, 22.76) 2.85 (1.92, 4.22) 0.37 (0.17, 0.85) 2.40 (1.44, 3.99)
6.61 3.99 7.32 7.80 7.39 7.41 4.75 9.50 6.84 61.61
0.45 (0.02, 8.75) 1.42 (0.75, 2.67) 2.28 (1.07, 4.83) 1.25 (0.57, 2.75) 1.49 (0.28, 7.82) 5.35 (1.46, 19.59) 1.72 (1.16, 2.56)
1.26 8.00 7.23 6.99 3.19 4.39 31.05
1.17 (0.56, 2.45) 1.17 (0.56, 2.45)
7.34 7.34
2.06 (1.45, 2.94)
100.00
%
ID
OR (95% CI)
Caucasian Hohler (1997)
8.17 (3.23, 20.67) 13.45
Reich (2002)
4.01 (2.39, 6.75)
Nedoszytko (2007)
8.79 (2.61, 29.68) 10.37
Subtotal (I−squared = 24.3%, P = 0.267)
5.52 (3.23, 9.43)
1
18.79 42.62
. Asian Li (2002)
2.32 (1.19, 4.51)
16.81
Zhang (2002)
2.88 (1.32, 6.31)
15.26
Long (2004)
8.78 (2.33, 33.03) 9.41
Subtotal (I−squared = 36.1%, P = 0.209)
3.19 (1.71, 5.96)
41.48
Magalhaes (2010)
1.17 (0.56, 2.45)
15.90
Subtotal (I−squared = .%, P = .)
1.17 (0.56, 2.45)
15.90
3.68 (2.17, 6.24)
100.00
. Latino
. Overall (I−squared = 66.0%, P = 0.007)
0.0234
Weight
0.0303
1
33
(d)
42.7
Study ID
(c)
OR (95% CI)
% Weight
Caucasian
Study ID
OR (95% CI)
% Weight
Asian
Hohler (1997)
8.17 (3.23, 20.67) 11.20
Mossner (2005)
3.21 (1.65, 6.23)
Nedoszytko (2007)
6.78 (2.02, 22.76) 9.43
Reich (2007)
3.64 (2.39, 5.55)
14.19
Gallo (2012)
0.37 (0.17, 0.85)
11.92
Subtotal (I−squared = 87.6%, P = 0.000)
2.91 (1.14, 7.42)
59.62
Zhang (2002)
12.87
Li (2002)
0.28 (0.06, 1.24)
37.48
Zhang (2002)
0.81 (0.17, 3.86)
14.59
Long (2004)
0.94 (0.09, 9.42)
5.91
Subtotal (I−squared = 0.0%, P = 0.530)
0.48 (0.19, 1.26)
57.98
2.28 (1.07, 4.83)
12.32
Kim (2003)
1.25 (0.57, 2.75)
12.07
Reich (2002)
1.87 (0.89, 3.94)
34.53
Tsunemi (2003)
1.49 (0.28, 7.82)
7.07
Nedoszytko (2007)
0.67 (0.07, 6.68)
7.49
Long (2004)
5.35 (1.46, 19.59) 8.93
42.02
Subtotal (I−squared = 21.6%, P = 0.281)
2.02 (1.15, 3.57)
40.38
2.55 (1.40, 4.67)
100.00
. Asian
. Caucasian
Subtotal (I−squared = 0.0%, P = 0.402)
1.66 (0.82, 3.36)
.
. Overall (I−squared = 31.2%, P = 0.214)
0.98 (0.56, 1.70)
0.0636
1
Overall (I−squared = 78.9%, P = 0.000)
100.00
0.0439
15.7
4
% OR (95% CI)
Weight
Hohler (1997)
6.26 (2.46, 15.94)
5.22
Hamamoto (2000)
0.45 (0.02, 8.75)
2.42
Gonzalez (2001)
3.63 (0.89, 14.77)
3.09
AI−Heresh (2002)
1.54 (0.74, 3.22)
16.68
Rahman (2006)
2.41 (1.17, 5.00)
15.63
Rahman (2006)
1.75 (0.85, 3.60)
17.19
Reich (2007)
2.13 (1.37, 3.32)
39.79
Overall (I−squared = 22.5%, P = 0.258)
2.23 (1.69, 2.95)
100.00
2
logOR
ID
22.8
Begg’s funnel plot with pseudo 95% confidence limits
(f)
(e) Study
1
0
–2 0.0234
1
42.7
0
0.5
1
1.5
s.e. of: logOR
Figure 1 Meta-analysis of the tumour necrosis factor (TNF)-a 238G/A polymorphism (rs361525) with the dominant model (AA/GA vs. GG). (a–e) Forest plots for (a) psoriasis overall with stratification for ethnic origin; (b) psoriasis type I (early-onset); (c) psoriasis type II (late-onset); (d) psoriasis vulgaris (PV); and (e) psoriatic arthritis (PsA). (f) Begg funnel plot for psoriasis overall.
0.63, I2 = 0.0%) (Fig. 2d). However, there was no association found with PsA (OR = 0.85, 95% CI = 0.69–1.04, POR = 0.12; v² = 5.30, Phet = 0.51,
840
Clinical and Experimental Dermatology (2013) 38, pp836–844
I2 = 0.0%) (Fig. 2e), and omission of the study by AIHeresh et al.16 changed the results (the pooled OR was 0.79 with 95% CI = 0.63–0.98).
ª 2013 British Association of Dermatologists
TNF-a polymorphism and psoriasis Y. Jia et al.
(b)
(a) Study ID
OR (95% CI)
Caucasian Hohler (1997) Craven (2001) Gonzalez (2001) AI−Heresh (2002) Mossner (2005) Baran (2006) Rahman (2006) Rahman (2006) Nedoszytko (2007) Reich (2007) Settin (2009) Magalhaes (2010) Gallo (2012) Subtotal (I−squared = 22.4%, P = 0.217) . Asian Hamamoto (2000) Li (2002) Zhang (2002) Kim (2003) Tsunemi (2003) Long (2004) Subtotal (I−squared = 0.0%, P = 0.737) . Overall (I−squared = 8.7%, P = 0.349)
0.65 (0.34, 1.23) 0.57 (0.27, 1.19) 0.70 (0.27, 1.81) 1.35 (0.77, 2.37) 0.51 (0.31, 0.83) 0.87 (0.40, 1.87) 1.01 (0.62, 1.64) 0.88 (0.53, 1.48) 0.43 (0.22, 0.85) 0.66 (0.50, 0.88) 0.27 (0.09, 0.81) 0.66 (0.29, 1.50) 0.75 (0.34, 1.66) 0.71 (0.61, 0.83)
0.0234
1
% Weight
Study ID
0.37 8.55 2.51 3.32 1.09 2.22 18.06
0.68 (0.59, 0.79)
100.00
42.7
% Weight
Caucasian
5.11 4.14 2.32 4.62 9.24 3.04 7.03 6.65 5.09 26.06 2.52 3.06 3.07 81.94
0.45 (0.02, 8.75) 0.57 (0.34, 0.96) 0.91 (0.39, 2.11) 0.43 (0.17, 1.08) 0.29 (0.05, 1.59) 0.35 (0.11, 1.16) 0.55 (0.38, 0.79)
OR (95% CI)
Hohler (1997)
0.43 (0.18, 1.03)
10.91
Reich (1999)
0.54 (0.30, 0.97)
19.49
Craven (2001)
0.46 (0.19, 1.12)
Baran (2006)
0.67 (0.27, 1.65)
7.67
Nedoszytko (2007)
0.42 (0.21, 0.87)
13.96
Magalhaes (2010)
0.66 (0.29, 1.50)
8.94
Subtotal (I−squared = 0.0%, P = 0.947)
0.52 (0.38, 0.71)
70.62
Li (2002)
9.65
. Asian 0.60 (0.33, 1.09)
18.43
Zhang (2002)
1.13 (0.47, 2.71)
5.96
Long (2004)
0.28 (0.06, 1.32)
4.99
Subtotal (I−squared = 25.9%, P = 0.259)
0.65 (0.41, 1.04)
29.38
0.56 (0.43, 0.72)
100.00
. Overall (I−squared = 0.0%, P = 0.794)
0.0595
1
16.8
(d) Study
(c)
%
ID
Study ID
OR (95% CI)
% Weight
OR (95% CI)
Weight
Caucasian
Caucasian
Hohler (1997)
0.43 (0.18, 1.03)
6.72
Mossner (2005)
0.51 (0.31, 0.83)
16.65
Baran (2006)
0.87 (0.40, 1.87)
5.48
Reich (1999)
0.75 (0.37, 1.52)
24.14
Nedoszytko (2007)
0.43 (0.22, 0.85)
9.16
Craven (2001)
0.75 (0.30, 1.88)
14.14
Reich (2007)
0.65 (0.47, 0.90)
35.45
Baran (2006)
1.38 (0.49, 3.88)
7.82
Settin (2009)
0.27 (0.09, 0.81)
Nedoszytko (2007)
0.45 (0.15, 1.34)
14.03
Gallo (2012)
0.75 (0.34, 1.66)
5.53
60.13
Subtotal (I−squared = 0.0%, P = 0.512)
0.58 (0.47, 0.72)
83.53
Zhang (2002)
0.91 (0.39, 2.11)
4.53
Kim (2003)
0.43 (0.17, 1.08)
5.99
Tsunemi (2003)
0.29 (0.05, 1.59)
1.96
Subtotal (I−squared = 0.0%, P = 0.536)
0.76 (0.49, 1.19)
4.54
.
.
Asian
Asian Zhang (2002)
0.33 (0.04, 2.65)
5.94
Li (2002)
0.54 (0.25, 1.13)
26.83
Long (2004)
0.48 (0.10, 2.30)
7.10
Long (2004)
0.35 (0.11, 1.16)
3.99
39.87
Subtotal (I−squared = 0.0%, P = 0.439)
0.53 (0.31, 0.88)
16.47
0.57 (0.47, 0.70)
100.00
Subtotal (I−squared = 0.0%, P = 0.912)
0.50 (0.26, 0.94)
.
. Overall (I−squared = 0.0%, P = 0.741)
0.66 (0.46, 0.94)
0.0419
1
Overall (I−squared = 0.0%, P = 0.626)
100.00
0.0513
23.9
4
% OR (95% CI)
Weight
Hohler (1997)
0.90 (0.43, 1.87)
7.63
Hamamoto (2000)
0.45 (0.02, 8.75)
0.85
Gonzalez (2001)
0.70 (0.27, 1.81)
5.31
AI−Heresh (2002)
1.35 (0.77, 2.37)
10.59
Rahman (2006)
1.01 (0.62, 1.64)
16.11
Rahman (2006)
0.88 (0.53, 1.48)
15.24
Reich (2007)
0.68 (0.49, 0.94)
44.26
Overall (I−squared = 0.0%, P = 0.506)
0.85 (0.69, 1.04)
100.00
2
logOR
ID
19.5
Begg’s funnel plot with pseudo 95% confidence limits
(f)
(e) Study
1
0
–2
–4 0.0234
1
42.7
0
0.5
1
1.5
s.e. of: logOR
Figure 2 Meta-analysis of the tumour necrosis factor (TNF)-a 308G/A polymorphism (rs1800629) with the dominant model (AA/GA vs. GG). (a–e) Forest plots for (a) psoriasis overall with stratification for ethnic origin; (b) psoriasis type I (early-onset); (c) psoriasis type II (late-onset); (d) psoriasis vulgaris (PV); and (e) psoriatic arthritis (PsA). (f) Begg funnel plot for psoriasis overall.
Sensitivity analyses found no significant influence of any one study to the incidence of psoriasis overall, psoriasis type I or PV. No evidence of publication bias
ª 2013 British Association of Dermatologists
was found for the overall study (P = 0.23 for Begg test and P = 0.22 for Egger test). The Begg funnel plot is shown in Fig. 2f.
Clinical and Experimental Dermatology (2013) 38, pp836–844
841
TNF-a polymorphism and psoriasis Y. Jia et al.
Discussion The present meta-analysis identified that carriers of the A allele (AA/GA) of the TNF-a 238G/A polymorphism had an increased risk of psoriasis overall, and that there were no significant differences between white and East Asian populations or between PV and PsA. However, when we divided psoriasis into two groups based on age of onset, we found that this increased risk was present only in psoriasis type I but not in type II, which was not consistent with the results of Li et al.4 It suggests that the susceptibility difference of the TNF-a 238G/A variant may be associated with age of onset only and not with classifications of either clinical features or ethnic origin. In addition, the positive associations of the A allele with psoriasis overall, as well as with PV and PsA, are possibly due to the large numbers of patients with type I psoriasis, who made up almost three-quarters of all patients. There was obvious heterogeneity between studies for psoriasis overall, psoriasis type I and PV, which was mainly related to the frequencies of the A and G alleles and the GA and GG genotypes in populations. Sensitivity analyses indicated that our results for the 238G/A polymorphism were statistically reliable. The meta-analysis of the 308G/A polymorphism showed that carriage of the A allele conferred a decreased risk for psoriasis in the overall study and each group except for PsA, and no heterogeneity between studies was found. However, the results for the type II psoriasis and PsA groups were not statistically reliable because the pooled OR had a significant difference, due to some study omissions. The morbidity rate of PV is much higher than that of PsA; however, in our meta-analysis, there was no significant difference between PV and PsA for the association of the TNF-a 238G/A polymorphism with psoriasis risk. Thus, the differences in morbidity rate between these two conditions may be attributable to other factors. For example, the TNF-857C/T polymorphism has been considered to be a genetic factor independent of the PSORS1 risk gene, which can influence the development of PsA.3 How the different roles of TNF-a –238 and –308 variants affect psoriasis risk remains unknown. The risk most probably relates to the variant-mediated abnormal gene expression;4 however, the functional effects of the two variants are still inconsistent.2,5,24 The human leucocyte antigen (HLA) region exhibits
842
Clinical and Experimental Dermatology (2013) 38, pp836–844
extensive linkage disequilibrium across the whole region, and TNF-a is known to be in strong linkage disequilibrium with some HLA haplotypes.6,15 Hence, it is difficult to tell whether the relationship between the polymorphisms and psoriasis is caused by the direct influence of the polymorphisms themselves or by linkage disequilibrium with other HLA genes. Inevitably, the data included in the present metaanalysis had some limitations. First, some controls were not in HWE, although this did not contribute to the heterogeneity seen in this meta-analysis. Second, some studies were not included in this meta-analysis because of the full text was unavailable or there were insufficient data, and some of the included studies were not adequate, especially in the subgroups. This probably led to an inaccurate estimate of publication bias, even though the relative statistical tests did not show this because the sensitivities of both the Begg and Egger tests were generally low in this meta-analysis, based on < 20 trials.25
Conclusion This meta-analysis supported an overall increased psoriasis risk associated with the TNF-a 238G/A polymorphism and a reduced risk with the 308G/ A polymorphism. The 238G/A polymorphism only had an effect in early-onset patients, and there were no significant differences in ethnicity or classifications of clinical features. For 308G/A, there were also no significant differences between white and East Asian populations; however, the accuracy and reliability of the results, especially in psoriasis type II and PsA, should be confirmed by well-designed studies with a larger sample size and disease classifications. Further, relative functional studies also should be performed to better explore the complex and comprehensive roles of the TNF-a promoter polymorphism.
What’s already known about this topic?
•
TNF-a may be involved in the aetiology of psoriasis. • The association between TNF-a -238G/A and 308G/A polymorphisms and psoriasis risk is still ambiguous.
ª 2013 British Association of Dermatologists
TNF-a polymorphism and psoriasis Y. Jia et al.
What does this study add?
•
There is an an overall increased psoriasis risk was associated with the TNF-a 238G/A polymorphism. • There is a reduced psoriasis risk with the TNFa 308G/A polymorphism.
References 1 Elder JT, Bruce AT, Gudjonsson JE et al. Molecular dissection of psoriasis: integrating genetics and biology. J Invest Dermatol 2010; 130: 1213–1226. 2 Elahi MM, Asotra K, Matata BM, Mastana SS. Tumor necrosis factor alpha -308 gene locus promoter polymorphism. An analysis of association with health and disease. Biochim Biophys Acta 2009; 1792: 163–172. 3 Reich K, Huffmeier U, Konig IR et al. TNF polymorphisms in psoriasis. Association of psoriatic arthritis with the promoter polymorphism TNF*-857 independent of the PSORS1 risk allele. Arthritis Rheum 2007; 56: 2056–2064. 4 Li C, Wang G, Gao Y et al. TNF-alpha gene promoter 238G>A and 308G>A polymorphisms alter risk of psoriasis vulgaris: a meta-analysis. J Invest Dermatol 2007; 127: 1886–1892. 5 Reich K, Mossner R, Konig IR et al. Promoter polymorphisms of the genes encoding tumor necrosis factor-alpha and interleukin-1beta are associated with different subtypes of psoriasis characterized by early and late disease onset. J Invest Dermatol 2002; 118: 155–163. 6 Hohler T, Kruger A, Schneider PM et al. A TNF-alpha promoter polymorphism is associated with juvenile onset psoriasis and psoriatic arthritis. J Invest Dermatol 1997; 109: 562–565. 7 Tsunemi Y, Nishibu A, Saeki H et al. Lack of association between the promoter polymorphisms at positions 308 and 238 of the tumor necrosis factor alpha gene and psoriasis vulgaris in Japanese patients. Dermatology 2003; 207: 371–374. 8 Long F, Sun C, Deng D et al. TNF-238A is associated with juvenile onset psoriasis in patients of Han population in southwest China. J Dermatol Sci 2004; 36: 109–111. 9 Mossner R, Kingo K, Kleensang A et al. Association of TNF 238 and 308 promoter polymorphisms with psoriasis vulgaris and psoriatic arthritis but not with pustulosis palmoplantaris. J Invest Dermatol 2005; 124: 282–284. 10 Rahman P, Siannis F, Butt C et al. TNF-alpha polymorphisms and risk of psoriatic arthritis. Ann Rheum Dis 2006; 65: 919–923.
ª 2013 British Association of Dermatologists
11 Gallo E, Cabaleiro T, Rom an M et al. Study of genetic polymorphisms in the tumor necrosis factor a promoter region in Spanish patients with psoriasis. Actas Dermosifiliogr 2012; 103: 301–307. 12 Reich K, Westphal G, Schulz T et al. Combined analysis of polymorphisms of the tumor necrosis factor-alpha and interleukin-10 promoter regions and polymorphic xenobiotic metabolizing enzymes in psoriasis. J Invest Dermatol 1999; 113: 214–220. 13 Hamamoto Y, Tateno H, Ishida T, Muto M. Lack of association between promoter polymorphism of the tumor necrosis factor-alpha gene and psoriatic arthritis in Japanese patients. J Invest Dermatol 2000; 115: 1162–1164. 14 Craven NM, Jackson CW, Kirby B et al. Cytokine gene polymorphisms in psoriasis. Br J Dermatol 2001; 144: 849–853. 15 Gonzalez S, Brautbar C, Martinez-Borra J et al. Polymorphism in MICA rather than HLA-B/C genes is associated with psoriatic arthritis in the Jewish population. Hum Immunol 2001; 62: 632–638. 16 Al-Heresh AM, Proctor J, Jones SM et al. Tumour necrosis factor-alpha polymorphism and the HLACw*0602 allele in psoriatic arthritis. Rheumatology (Oxford) 2002; 41: 525–530. 17 Li B. Tumor Necrosis Factor-Α Gene Polymorphism and Psoriasis. PhD Thesis. Shenyang: China Medical University, 2002. 18 Zhang AP. Experimental Studies on HLA Genotyping in Patients with Psoriasis Vulgaris in Chinese Han Population. PhD Thesis. Nanjing: Nanjing Medical University, 2002. 19 Kim TG, Pyo CW, Hur SS et al. Polymorphisms of tumor necrosis factor (TNF) alpha and beta genes in Korean patients with psoriasis. Arch Dermatol Res 2003; 295: 8–13. 20 Baran W, Szepietowski JC, Mazur G, Baran E. A 308 promoter polymorphism of tumor necrosis factor alpha gene does not associate with the susceptibility to psoriasis vulgaris. No difference either between psoriasis type I and type II patients. Acta Dermatovenerol Alp Panonica Adriat 2006; 15: 113–118. 21 Nedoszytko B, Szczerkowska-Dobosz A, Zablotna M et al. Associations of promoter region polymorphisms in the tumour necrosis factor-alpha gene and early-onset psoriasis vulgaris in a northern Polish population. Br J Dermatol 2007; 157: 165–167. 22 Settin A, Hassan H, El-Baz R, Hassan T. Association of cytokine gene polymorphisms with psoriasis in cases from the Nile Delta of Egypt. Acta Dermatovenerol Alp Panonica Adriat 2009; 18: 105–112. 23 Magalhaes RF, Biral AC, Pancoto JA et al. Human leukocyte antigen (HLA) and single nucleotide polymorphisms (SNPs) tumor necrosis factor (TNF)-alpha -238 and -308 as genetic markers of susceptibility to
Clinical and Experimental Dermatology (2013) 38, pp836–844
843
TNF-a polymorphism and psoriasis Y. Jia et al.
psoriasis and severity of the disease in a long-term follow-up Brazilian study. Int J Dermatol 2010; 49: 1133–1140. 24 Kaluza W, Reuss E, Grossmann S et al. Different transcriptional activity and in vitro TNF-alpha production in psoriasis patients carrying the TNF-alpha
844
Clinical and Experimental Dermatology (2013) 38, pp836–844
238A promoter polymorphism. J Invest Dermatol 2000; 114: 1180–1183. 25 Sterne JA, Gavaghan D, Egger M. Publication and related bias in meta-analysis: power of statistical tests and prevalence in the literature. J Clin Epidemiol 2000; 53: 1119–1129.
ª 2013 British Association of Dermatologists
