Polymorphism in STAT4 Increase the Risk of Systemic Lupus Erythematosus: An Updated Meta-Analysis

Previous studies have reported that STAT4 rs7574865 conferred the susceptibility to systemic lupus erythematosus (SLE). In this study, a meta-analysis (including 32 comparative studies of 11384 patients and 17609 controls) was conducted to investigate the role of STAT4 polymorphism in SLE in a comprehensive way. We found that the Asian population had the highest prevalence of the T allele than any other study population at 32.2% and that STAT4 rs7574865 polymorphism was associated with SLE in the overall population (OR = 1.579, 95%CI = 1.497-1.665, P < 0.001). In the subgroup analysis by ethnicity, STAT4 rs7574865 T allele was shown to be risk factor in SLE in Asian, European, and American origins. Our results do support STAT4 rs7574865 polymorphism as a susceptibility factor for SLE in populations of different ethnic and that its prevalence is ethnicity dependent.


Introduction
Systemic lupus erythematosus (SLE) is classified as an inflammatory autoimmune disease characterized by production of autoantibodies to nuclear antigens and immune complex formation [1], leading to multiple organs and systems damage, such as lupus nephritis, lupus encephalopathy, interstitial lung disease, and damage of blood system. Although the etiology of SLE still remains not completely clear, it was generally known that genetic and environmental factors interacting with each other contribute to SLE risk [2].
Genome-wide association studies have identified signal transducer and activator of transcription4 (STAT4) as a novel susceptibility gene associated with SLE [3]. STAT4 consists of 24 exons that spread over a 120 kb region on chromosome 2q32.3, and T is the mutation allele. It encodes a transcription factor that transmits signals. The main STAT4 activating cytokines are IL-12 and IL-23, leading to T-helper type1 and T-helper type17 differentiation with IFN-? and IL-17 production, which are critical players in the pathogenesis of SLE [4,5].
As the investigation varied greatly among studies, the association between STAT4 rs7574865 polymorphism and SLE risk has reported mixed results , which may due to small sample sizes, leading to inadequate statistical power, or interstudy inconsistencies, such as ethnic differences. Therefore, to overcome the limitations of these individual studies and resolve the observed inconsistencies, an exhaustive meta-analysis was performed to reach a credible conclusion. relevant articles were retrieved by keyword combinations as follows: "STAT4," "Signal transducer and activator of transcription 4," "rs7574865," "polymorphism," and "systemic lupus erythematosus." The references in the obtained studies were also examined to identify additional studies not included by the electronic databases, with no language or country restrictions.
2.2. Criteria for Inclusion. The inclusion criteria were listed as following: case-control studies, original data (independence among studies), and enough data to calculate an odds ratio (OR) with 95% confidence interval (CI). We excluded the following: studies in which the number of risk alleles could not be ascertained or contained overlapping data. These studies were assessed by two investigators independently. In case of any disagreement, it could be resolved by consultation with a third author.

Data Extraction.
We extracted the following information from each article: first author, year of publication, demographics, ethnicity, the numbers of cases and controls, and allele frequency of the STAT4 rs7574865 polymorphism.

Statistical Analyses.
We performed meta-analyses to estimate the association between STAT4 rs7574865 polymorphisms and SLE risk by odds ratios (ORs) and its 95% confidence interval (CI). The significance of the pooled ORs was determined by the Z-test, and P < 0:5 was considered statistically significant. In addition, Cochran , s Q statistics was used to assess within-study and between-study variation and heterogeneities. This heterogeneity test assesses the null hypothesis that all studies evaluated the same effect. The effect of heterogeneity was quantified using I 2 . In our meta-analysis, P < 0:10 for the Q-test and I 2 more than 50% was considered that significant heterogenicity among studies could not be ignored. Data were combined using both the fixed effects model (the inverse varianceweighted method) and the random effects model (DerSimonian and Laird method) [27,28]. We carried out the random effects model when the effects are assumed heterogeneous in our study. All analyses were performed using the STATA 15.1 software.

Evaluation of Heterogeneity and Publication
Bias. Subgroup analyses were performed by ethnicity, to examine the potential sources of heterogeneity observed in the meta-analyses. Because of the limitations of funnel plotting, publication bias was evaluated using the Egger linear regression test [29], which measures funnel plot asymmetry using a natural logarithm scale of ORs. We carried out sensitivity analysis to evaluate the effect of individual study on pooled OR. "Trim and fill" method was used to adjust summary estimates for observed bias [30], when asymmetry was indicated. In this method, small studies were removed until symmetry is achieved in the funnel plot by recalculating the center of the funnel, and then removed studies are replaced with their missing mirror-image counterparts. Finally, the revised summary estimate was calculated using all the original and hypothetical "filled" studies. Figure 1 showed the literature searching process. 133 references were found by electronic and manual searches, of which 104 records were excluded by means of reading the title and abstract. Therefore, 29 full-text publications were judged potentially relevant, comprehensively assessed against inclusion criteria. Ultimately, 21 articles were included in our meta-analysis , excluding 8 studies with duplicate data and no case controls ( Figure 1). Of these, one study contained data from five different groups [26], three studies contained data from three different groups, respectively [11,15,25], and one study contained data from two different groups [23]. These groups were analyzed independently, of which a total of 32 separate comparisons were considered, including 11384 SLE patients and 17609 controls. The major characteristics of individual eligible article were listed in Table 1 and  3.2. Frequency of the STAT4 rs7574865 T Allele in Different Ethnic Groups. As shown in Table 2, Asian had the greatest prevalence of STAT4 rs7574865 T allele and Europeans with lower prevalence. The mean frequency of T allele was 29.5%, and it varied from 21.8% to 32.2% in controls.   The results of subgroup analysis revealed that betweenstudy heterogeneity was moderately obvious in European (I 2 = 53:3%, P < 0:05) and American (I 2 = 69:7%, P < 0:05), but no such heterogeneity was found in Asian (I 2 = 18:9%, P > 0:05). (Table 3, Figure 3).

Publication Bias and Sensitivity Analyses. We performed
Egger's regression model and funnel plot to explore the potential publication bias of literatures, which results revealed no evidence of publication bias in individual groups (Egger regression P value > 0.05). (Table 3, Figure 4). Sensitivity analysis was conducted to evaluate the influence of a single study on pooled OR by the "trim and fill" method. Corresponding pooled OR was not substantially changed when any individual study was deleted, which confirmed the stability of our overall result. The estimated 3 International Journal of Rheumatology number of missing studies is 5, and the adjusted pooled OR estimated is 1.514 (95%CI = 1:462-1.567), which showed still significantly increased risk for SLE ( Figure 5).

Discussion
Systemic lupus erythematosus is a highly heterogeneous autoimmune disease, with influence of complex genetic. Studies in multiple ethnic populations have reported up to 100 risk loci of importance in SLE, although the mechanisms underlying these associations are still elusive [31][32][33].
Previous studies suggest that several single-nucleotide polymorphisms (SNPs) in LD in the third intron of STAT4, tagged by rs7574865, were initially demonstrated as SLE risk variants [25]. The minor T allele of rs7574865 is strongly linked with SLE [16]. Moreover, the STAT4 risk allele is also associated with specific clinical manifestations including earlier age at diagnosis, presence of anti-dsDNA, ischemic cerebrovascular disease, nephritis, and severe renal insufficiency [24,[34][35][36]. The above phenomena were consistent with the recent study which highlighted the role of STAT4 polymorphism in susceptibility to SLE [37]. However, due to the heterogeneity among different ethnic groups and a small sample size, existing evidences were not consistent.
In our comprehensive meta-analysis, we identified 21 relevant studies including 11384 SLE cases and 17609 healthy controls. The results do support a trend of association between STAT4 rs7574865 polymorphism and SLE risk in overall. When stratified by ethnicity for STAT4 rs7574865 polymorphism, positive correlations were observed in Asian, European, or American origin. Regarding STAT4 rs7574865 polymorphism in overall samples, we observed that the average frequency of T allele was 39.4% in patients with SLE compared with 29.5% in controls. The prevalence of the rs7574865 T allele was found to vary among ethnic populations.   Nevertheless, several limitations presented in this metaanalysis should be considered. First of all, the association between the STAT4rs7574865 polymorphism and disease severity and/or clinical features could not be performed limitation of amount of available data. Secondly, our ethnicity-associated results are applicable only to Asian, European, and American patients, as the data obtained from these ethnic groups. In addition, the heterogeneity has been presented obviously in European and American population, in subgroup analysis. Due to population demographics and lower  5 International Journal of Rheumatology SLE risk predisposition, or publication bias, fewer relevant studies were obtained in European and American, which may have distorted the heterogeneity, even have affected the final result, despite performing the "trim and fill" method and Egger regression test. Last but not least, STAT4 polymorphism may interact with other potential confounding risk factors.
In summary, our results stressed the importance of STAT4rs7574865 polymorphisms in increasing the risk of SLE in multiple ethnic groups and that the prevalence of the STAT4rs7574865 T allele may depend on ethnicity, which may improve our understanding of SLE pathogenesis. However, additional research with well-designed and large sample sizes is needed to be carried out about European, American, and other ethnicities in the future.

Data Availability
The extracted data used to support the findings of this study are mainly included within the article.

Conflicts of Interest
The authors declare that they have no conflicts of interest.