Association Between TNF-α -308G/A Polymorphism and Risk of Immune Thrombocytopenia: A Meta-Analysis

Abstract

Objective:

Previous studies have investigated the association between tumor necrosis factor-alpha (TNF-α) -308G/A polymorphism and risk of immune thrombocytopenia (ITP), but the reported results have been inconsistent. Thus, a systematic meta-analysis was performed to resolve this discrepancy.

Methods:

Electronic databases and the cited references of the obtained published articles were manually searched. Quality assessment of each study was conducted using the Newcastle-Ottawa Scale (NOS). All case-control studies were used to assess the strength of the association. Statistical analysis was performed using Stata version 12.0.

Results:

Eight high-quality studies, including 947 patients and 1911 controls, were selected for the final meta-analysis. There was no significant association between TNF-α -308G/A polymorphism and ITP in overall and Asian populations. However, a significant positive association was observed between them in the dominant genetic model (AA+AG versus GG) in the Caucasian population (OR = 1.35, 95% confidence interval [CI]: 1.07-1.71, P_H = 0.173).

Conclusions:

Our finding suggested that TNF-α -308G/A might be involved in development of ITP in the Caucasian population, but not in the Asian population. Among Caucasians the A allele (AA+AG) was associated with ITP. However, larger-scale studies are required to confirm our findings.

Introduction

Immune thrombocytopenia (ITP) is an acquired autoimmune disease characterized by autoantibody-dependent accelerated destruction and immature megakaryocytes leading to decreased platelet production (Kistangari and McCrae, 2013). Approximately 6.4 per 100,000 children and 3.3 per 100,000 adults were newly diagnosed per year (Terrell et al., 2010). Although its etiology remains unknown, both genetic and environmental factors are main causes leading to the development of this disease (Nadeem et al., 2015). Among the genetic factors, single-nucleotide polymorphisms (SNPs) in tumor necrosis factor-alpha (TNF-α) have been examined as candidate factors that might play a role in susceptibility of ITP (Foster et al., 2001; Satoh et al., 2004; Rocha et al., 2010; Okulu et al., 2011; Pehlivan et al., 2011; Sarpatwari et al., 2011; El Sissy et al., 2014; Wang et al., 2014; Pavkovic et al., 2015; Yadav et al., 2016).

TNF-α is a potential proinflammatory cytokine that plays a prominent role in the inflammatory and immune response, and it is involved in regulation of cell proliferation and differentiation (Qidwai and Khan, 2011). SNPs are common in the promoter region of TNF-α. Several polymorphisms such as -238A/G (rs361525), -308G/A (rs1800629), -857C/T (rs1799724), -863C/A (rs1800630), and -1031T/C (rs1799964) in the region were reported to influence the expression of TNF-α (Hellmig et al., 2005; Somi et al., 2006). Moreover, overexpression of TNF-α was observed in many inflammatory diseases, such as type 2 diabetes mellitus, allergic disease, and autoimmune disease (Schulz et al., 2014; Sefri et al., 2014; Li et al., 2015). Previous studies reported that TNF-α -308G/A could affect its transcription (Messer et al., 1991; Bidwell et al., 1999), and the locus was speculated to increase the risk and exacerbate the outcome of ITP.

Despite numerous previous studies that reported the relationship between TNF-α -308G/A and ITP in various populations and regions (Foster et al., 2001; Satoh et al., 2004; Rocha et al., 2010; Okulu et al., 2011; Pehlivan et al., 2011; Sarpatwari et al., 2011; El Sissy et al., 2014; Wang et al., 2014; Pavkovic et al., 2015; Yadav et al., 2016), the results were not consistent and conflicting. Some studies demonstrated no significant association between them (Satoh et al., 2004; Okulu et al., 2011; Wang et al., 2014; Yadav et al., 2016). However, others demonstrated that A allele of the locus was associated with an elevated cytokine production, contributing to increased risk of ITP (Foster et al., 2001; Pehlivan et al., 2011; Sarpatwari et al., 2011). Therefore, to better clarify the association, this meta-analysis aimed to synthesize data from different studies to elucidate whether TNF-α -308G/A polymorphism was associated with increased susceptibility to ITP.

Materials and Methods

Literature search

All case-control studies concerning association between TNF-α -308G/A and ITP were identified by searching relevant electronic databases of PubMed, Web of Science, Ovid, Scopus, Chinese Biomedical Literature Database, Wan Fang DATA, and China National Knowledge Infrastructure until March 2016. The following keywords were included: tumor necrosis factor OR TNF AND polymorphism OR variant OR variation OR mutation OR genotype AND ([immune thrombocytopenia] OR [idiopathic thrombocytopenic purpura] OR [immune thrombocytopenic purpura] OR [idiopathic thrombocytopenia] OR ITP). In addition, all relevant references mentioned in the full-text publications were manually searched to retrieve potential studies.

Inclusion and exclusion criteria

The inclusion criteria were included as follows: (1) ITP patients were confirmed according to the diagnostic criteria used by the International Working Group (Rodeghiero et al., 2009; Neunert et al., 2011); (2) the case-control studies evaluated association between TNF-α -308G/A and ITP; (3) sufficient information in cases and controls could be acquired to estimate odds ratio (OR) with 95% confidence interval (CI); and (4) the studies contained no restricted language. Meanwhile, duplicate study and a study without detail genotype data or the data not in agreement with Hardy-Weinberg equilibrium (HWE) were excluded in our study.

Data extraction

The data of the eligible studies were independently extracted by two investigators (Jing Zhang and Yan-Mei Xu). Any discrepancies were resolved after discussion, and a consensus was ultimately reached. The following information was extracted: name of the first author, year of publication, country, ethnicity, disease type, mean age, sample size, genotype distribution in cases and controls, and p-value of HWE in controls.

Quality assessment

The quality of each study was assessed independently by two investigators (Yan Wang and Fan Sun) in accordance with the Newcastle-Ottawa Scale (NOS), which awarded studies with a star scoring scale (Stang, 2010). A study can be given a maximum of one star for each numbered item within the Selection and Outcome. A maximum of two stars can be assigned for Comparability. High-quality studies are indentified with a NOS score of five or more, whereas those with less than a score of five were considered to be low-quality studies.

Statistical analyses

The association between TNF-α -308 G/A and ITP was assessed by pooled OR with 95% CI. Meta-analyses were performed for all genetic models. HWE in controls was evaluated by the chi-square test (Wang, 2012). Heterogeneity among studies was examined with the inconsistency index (I²) statistic (Higgins et al., 2003). If the value of I² was more than 50% and a p-value of heterogeneity test (P_H) was less than 0.10, the random effect model would be used to calculate the pooled OR. Otherwise, a fixed effects model was used to assess the heterogeneity (Zintzaras and Ioannidis, 2005). Sensitivity analysis was conducted to assess the effect of each study by excluding one study at a time. Publication bias was detected by Begg's funnel plot and Egger's regression test (Egger et al., 1997), and p < 0.05 was considered as a significant publication bias. All statistical analyses were performed with Stata 12.0 software (StataCorp, College Station, TX).

Results

Study characteristics

The search strategy initially identified 494 potentially relevant studies based on the inclusion criteria, and 55 duplicate publications were excluded. By screening the title and abstract, we deemed 413 articles to be incongruous with the aim of our study. Thus, 26 records remained, of which 15 were removed for not being related to the TNF-α -308G/A polymorphism and just one was removed for lack of providing the accurate genotype data. However, genotype distributions of the locus in two studies were not in accordance with HWE test (Foster et al., 2001; Satoh et al., 2004). At last, a total of eight case-control studies, including 947 cases and 1911 healthy controls, were included in the final meta-analysis (Rocha et al., 2010; Okulu et al., 2011; Pehlivan et al., 2011; Sarpatwari et al., 2011; El Sissy et al., 2014; Wang et al., 2014; Pavkovic et al., 2015; Yadav et al., 2016) (Fig. 1). Five eligible studies are described in Table 1, and the detail quality assessment of the case-control studies is listed in Supplementary Table S1 (Supplementary Data are available online at www.liebertpub.com/gtmb).

FIG. 1.

Flow diagram of study selection process.

Table 1.

Characteristics of Studies Included in This Meta-Analysis

						Case				Control						NOS
First author	Year	Age	Type	Ethnicity	Case/control	GG	AG	AA	A allele (%)	GG	AG	AA	A allele (%)	Polymorphisms	HWE (p)	Score
Yadav	2016	Mixed	Mixed	Asian	80/100	70	10	0	6.3	90	10	0	5.0	-308G/A	0.60	7
Pavkovic	2015	Adult	Chronic	Caucasian	125/120	105	19	1	8.4	95	23	2	11.3	-308G/A -308G/A,238G/A	0.66	8
Wang	2014	Adult	Acute	Asian	215/206	194	21	0	4.9	193	13	0	3.2	-857C/T, -863C/A, -1031T/A	0.64	8
El Sissy	2014	Pediatric	Acute	Caucasian	12/100	5	4	3	41.7	63	31	6	21.5	-308G/A	0.41	8
El Sissy	2014	Pediatric	Chronic	Caucasian	80/100	38	18	24	41.3	63	31	6	21.5	-308G/A	0.41	8
Sarpatwari	2011	Adult	NA	Caucasian	206/618	115	74	12	23.8	403	191	24	19.3	-308G/A	0.82	6
Pehlivan	2011	Adult	Chronic	Caucasian	71/71	56	15	0	10.6	64	7	0	4.9	-308G/A	0.66	7
Okulu	2011	Pediatric	Acute	Caucasian	25/38	20	5	0	10.0	37	11	0	14.5	-308G/A	0.37	8
Okulu	2011	Pediatric	Chronic	Caucasian	25/38	21	4	0	8.0	37	11	0	14.5	-308G/A	0.37	8
Rocha	2010	Adult	Chronic	South America	122/541	96	24	2	11.5	403	123	13	13.8	-308G/A	0.33	5
Satoh	2004	Adult	Chronic	Asian	84/56	80	4	0	2.4	56	0	0	0	-308G/A,238G/A	NA	8
Foster	2001	Pediatric	Chronic	Caucasian	37/218	33	3	1	6.8	142	57	0	13.1	-308G/A	0.02	4

HWE, Hardy-Weinberg equilibrium; NA, not available; NOS, Newcastle-Ottawa Scale.

Overall analysis

Overall, eight eligible studies were pooled in this study; no evidence of a significant association was observed between TNF-α -308G/A and ITP risk in any genetic model (dominant model: A/A+A/G versus G/G, OR = 1.11, 95% CI = 0.80-1.53, I² = 52.4%, P_H = 0.04; allele model: G versus A: OR = 0.87, 95% CI = 0.60-1.25, I² = 70.3%, P_H = 0.001; A/G versus G/G: OR = 0.87, 95% CI = 0.71-1.07, I² = 17.2%, P_H = 0.294) (Supplementary Figs. S1 and S2; Fig. 2). As some studies in our meta-analysis reported that the distribution of AA was absent among patients and controls (Okulu et al., 2011; Pehlivan et al., 2011; Wang et al., 2014; Yadav et al., 2016), data of the two genetic models (GG versus AA and AG+GG versus AA) were not available.

FIG. 2.

Forest plots for TNF-α -308G/A polymorphism and ITP risk in genetic model (AG vs. GG). ITP, immune thrombocytopenia; TNF-α, tumor necrosis factor-alpha.

Subgroup analysis

Three subgroups were stratified on the basis of ethnicity, mean age, and the type of ITP. There were no statistically significant differences in age and ITP type between cases and controls. However, ethnicity analysis revealed a significant association between them in the dominant model (AA+AG versus GG in Caucasians) (OR = 1.35, 95% CI 1.07-1.71, I² = 33.40%, P_H = 0.173), but not in Asians (OR = 0.81, 95% CI: 0.46-1.43, I² = 31.60%, P_H = 0.227) (Fig. 3).

FIG. 3.

Meta-analysis for the association between ITP risk and the TNF-α -308A/G polymorphism (dominant model: AA+AG vs. GG).

Sensitivity analysis

Sensitivity analysis was conducted to assess the effect of each study by omitting a single study at a time. After excluding the study deviation from HWE (Foster et al., 2001; Satoh et al., 2004), the results of the meta-analysis were stable.

Publication bias

Begg's funnel plot and Egger's regression test were selected to evaluate publication bias in our study. The funnel plot showed almost symmetrical (Fig. 4), and p-value of Egger's test (p = 0.92) was larger than 0.05. These results indicated that no potential publication bias was observed in our study.

FIG. 4.

Funnel plots for TNF-α -308G/A polymorphism and ITP risk (AG vs. GG).

Discussion

The present meta-analysis of eight high-quality case-control studies, including 947 patients and 1911 controls, revealed that TNF-α -308G/A under the dominant model (AA+AG versus GG) might significantly increase ITP risk in Caucasians, however, no statistical significant association was observed in overall and Asian populations, revealing that allele A carrier (AA+AG) of rs1800629 might increase predisposition to ITP in Caucasians.

The following reasons may explain our findings. Owing to differences in ethnic genetic backgrounds in Asians and Caucasians, frequency of the A allele is conspicuously higher in the Caucasian population than in the Asian population (Table 1). Moreover, only two studies, including a total of 295 patients and 306 controls, were conducted in the Asian population, and the relative small sample sizes of cases and controls might limit the power to reach a more reliable result. Besides, TNF-α -308G/A polymorphism was reported to affect gene transcription by increasing TNF-α production and regulating cell proliferation and differentiation, leading to potentially exacerbate the outcomes of ITP (Qidwai and Khan, 2011).

In addition, we should emphasize the importance of the heterogeneity and publication bias, which could affect the results of meta-analysis (Sterne and Egger, 2001). Supplementary Figure S3 shows that the study with low quality implemented by Foster et al. (2001) could cause publication bias. Table 1 presents that the data detected by Satoh et al. (2004) and Foster et al. (2001) were not in agreement with HWE. Once the two studies were excluded, the funnel plot showed almost symmetrical, suggesting deviation of allelic distribution from HWE might influence publication bias. Moreover, there was statistically significant heterogeneity between them in the dominant genetic model (Table 2). After the ethnicity analysis, the heterogeneity could effectively decrease, indicating that the ethnicity analysis, to some extent, may be used to explore the source of heterogeneity. Interestingly, analysis of frequency of A allele revealed that the prevalence of A allele was considerably lower in Asians than in Caucasians among cases and healthy controls (Table 1), suggesting that the rarity of the A allele hardly exhibits polymorphism in Asians. Therefore, the distributional difference of the A allele may contribute to the probability of false results.

Table 2.

Summary of Meta-Analysis Results in Each Subgroup

			Test of association			Test of heterogeneity
Genetic model	Group	Number of studies	OR	95% CI	p	model	p	I² (%)
AA+AG vs GG	Overall	8	1.11	0.80-1.53	0.543	R	0.040	52.40
	Caucasian	7	1.35	1.07-1.71	0.012^*	F	0.173	33.40
	Asian	2	0.81	0.46-1.43	0.472	F	0.227	31.60
	Chronic	5	1.22	0.75-2.00	0.419	R	0.054	57.00
	Acute	3	1.07	0.48-2.36	0.872	F	0.161	45.30
	Adult	5	1.01	0.67-1.53	0.959	R	0.034	61.70
	Pediatric	4	1.42	0.82-2.47	0.207	F	0.288	20.30
GG vs AG	Overall	8	0.87	0.71-1.07	0.196	F	0.294	17.20
	Caucasian	7	1.19	0.93-1.52	0.177	F	0.382	5.90
	Asian	2	1.48	0.84-2.06	0.176	F	0.710	0.00
	Chronic	5	1.08	0.80-1.48	0.610	F	0.302	17.70
	Acute	3	0.72	0.41-1.25	0.246	F	0.640	0.00
	Adult	5	1.18	0.94-1.49	0.157	F	0.113	46.50
	Pediatric	4	0.94	0.57-1.56	0.803	F	0.803	0.00
G vs A	Overall	8	0.87	0.60-1.25	0.452	R	0.001	70.30
	Caucasian	7	0.70	0.46-1.05	0.083	R	0.012	63.30
	Asian	2	1.18	0.61-2.31	0.620	F	0.237	28.60
	Chronic	5	0.85	0.44-1.62	0.615	R	0.001	79.40
	Acute	3	0.87	0.38-1.99	0.739	R	0.057	65.10
	Adult	5	1.00	0.68-1.47	0.989	R	0.034	61.70
	Pediatric	4	0.61	0.32-1.18	0.140	F	0.288	20.30

Data for p < 0.05.

F, fixed effects models; R, random effects model.

Our meta-analysis is the first assessment of the relationship between TNF-α -308G/A and risk of ITP. However, some limitations of this meta-analysis should be addressed. First of all, the number of studies and sample size included in this meta-analysis are relatively small. In addition, some eligible negative studies may not be published. Consequently, some potential publication bias may occur inevitably in our results. Finally, owing to insufficient data, we could not carry out the analysis of gene-environment interaction in this meta-analysis.

In conclusion, TNF-α -308G/A polymorphism may play a role in development of ITP, and allele A carrier (AA+AG) of the locus was a possible susceptible factor for Caucasians. Considering the limitations of the current study, a larger sample size in Asians and a potential target for therapeutic intervention in ITP are warranted to further validate the results.

Footnotes

Acknowledgment

This report was supported by the National Natural Science Foundation of China (No. 81271912 and No. 81360083).

Author Disclosure Statement

No competing financial interests exist.

References

Bidwell

, Wood

, Morse

, et al. (1999) Human cytokine gene nucleotide sequence alignments: supplement 1. Eur J Immunogenet, 26:135-223.

Egger

, Smith

, Schneider

, et al. (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ, 315:629-634.

El Sissy

, El Sissy

, Elanwary

(2014) Tumor necrosis factor-alpha -308G/A gene polymorphism in Egyptian children with immune thrombocytopenic purpura. Blood Coagul Fibrinolysis, 25:458-463.

Foster

, Zhu

, Erichsen

, et al. (2001) Polymorphisms in inflammatory cytokines and Fcgamma receptors in childhood chronic immune thrombocytopenic purpura: a pilot study. Br J Haematol, 113:596-599.

Hellmig

, Fischbach

, Goebeler-Kolve

, et al. (2005) A functional promotor polymorphism of TNF-alpha is associated with primary gastric B-Cell lymphoma. Am J Gastroenterol, 100:2644-2649.

Higgins

JPT

, Thompson

, Deeks

, et al. (2003) Measuring inconsistency in meta-analyses. BMJ, 327:557-560.

Kistangari

, McCrae

(2013) Immune thrombocytopenia. Hematol Oncol Clin North Am, 27:495-520.

Messer

, Spengler

, Jung

, et al. (1991) Polymorphic structure of the tumor necrosis factor (TNF) locus: an NcoI polymorphism in the first intron of the human TNF-beta gene correlates with a variant amino acid in position 26 and a reduced level of TNF-beta production. J Exp Med, 173:209-219.

Nadeem

, Mumtaz

, Naveed

, et al. (2015) Gene-gene, gene-environment, gene-nutrient interactions and single nucleotide polymorphisms of inflammatory cytokines. World J Diabetes, 6:642-647.

10.

Neunert

, Lim

, Crowther

, et al. (2011) The American Society of Hematology 2011 evidence-based practice guideline for immune thrombocytopenia. Blood, 117:4190-4207.

11.

Okulu

, Ileri

, Culha

, et al. (2011) The role of tumor necrosis factor-alpha-308 G/A and transforming growth factor-beta 1-915 G/C polymorphisms in childhood idiopathic thrombocytopenic purpura. Turk J Hematol, 28:170-175.

12.

Pavkovic

, Angelovic

, Popova-Simjanovska

, et al. (2015) Single nucleotide polymorphisms of the inflamatory cytokine genes: interleukin-1B, tumor necrosis factors-A and tumor necrosis factor-B in adult patients with immune thrombocytopenia. Prilozi, 36:109-115.

13.

Pehlivan

, Okan

, Sever

, et al. (2011) Investigation of TNF-alpha, TGF-beta 1, IL-10, IL-6, IFN-gamma, MBL, GPIA, and IL1A gene polymorphisms in patients with idiopathic thrombocytopenic purpura. Platelets, 22:588-595.

14.

Qidwai

, Khan

(2011) Tumour necrosis factor gene polymorphism and disease prevalence. Scand J Immunol, 74:522-547.

15.

Rocha

, De Souza

, Rocha

, et al. (2010) IL1RN VNTR and IL2-330 polymorphic genes are independently associated with chronic immune thrombocytopenia. Br J Haematol, 150:679-684.

16.

Rodeghiero

, Stasi

, Gernsheimer

, et al. (2009) Standardization of terminology, definitions and outcome criteria in immune thrombocytopenic purpura of adults and children: report from an international working group. Blood, 113:2386-2393.

17.

Sarpatwari

, Bussel

, Ahmed

, et al. (2011) Single nucleotide polymorphism (SNP) analysis demonstrates a significant association of tumour necrosis factor-alpha (TNFA) with primary immune thrombocytopenia among Caucasian adults. Hematology, 16:243-248.

18.

Satoh

, Pandey

, Okazaki

, et al. (2004) Single nucleotide polymorphisms of the inflammatory cytokine genes in adults with chronic immune thrombocytopenic purpura. Br J Haematol, 124:796-801.

19.

Schulz

, Reichert

, Streetz

, et al. (2014) Tumor necrosis factor-alpha and oral inflammation in patients with Crohn disease. J Periodontol, 85:1424-1431.

20.

Sefri

, Benrahma

, Charoute

, et al. (2014) TNF A −308G>A polymorphism in Moroccan patients with type 2 diabetes mellitus: a case-control study and meta-analysis. Mol Biol Rep, 41:5805-5811.

21.

Somi

, Najafi

, Noori

, et al. (2006) Tumor necrosis factor-alpha gene promoter polymorphism in Iranian patients with chronic hepatitis B. Indian J Gastroenterol, 25:14-15.

22.

Stang

(2010) Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol, 25:603-605.

23.

Sterne

JAC

, Egger

(2001) Funnel plots for detecting bias in meta-analysis: guidelines on choice of axis. J Clin Epidemiol, 54:1046-1055.

24.

Terrell

, Beebe

, Vesely

, et al. (2010) The incidence of immune thrombocytopenic purpura in children and adults: a critical review of published reports. Am J Hematol, 85:174-180.

25.

Wang

(2012) Statistical tests of genetic association for case-control study designs. Biostatistics, 13:724-733.

26.

Wang

, Li

, Wang

, et al. (2014) TNF-alpha promoter single nucleotide polymorphisms and haplotypes associate with susceptibility of immune thrombocytopenia in Chinese adults. Hum Immunol, 75:980-985.

27.

Yadav

, Tripathi

, Kumar

, et al. (2016) Association of TNF-alpha -308G>A and TNF-beta +252A>G genes polymorphisms with primary immune thrombocytopenia: a North Indian study. Blood Coagul Fibrinolysis, 27:791-796.

28.

Zintzaras

, Ioannidis

JPA

(2005) Heterogeneity testing in meta-analysis of genome searches. Genet Epidemiol, 28:123-137.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.06 MB

0.05 MB

0.03 MB

0.02 MB