RNF213 polymorphism and Moyamoya disease: A systematic review and meta-analysis
Correspondence Address: Source of Support: This article is supported by the Youth Project No.
30801185 (to Y.L.) and the General Program No.30872673 (to C.Y.)
from the National Natural Science Foundation of China, Conflict of Interest: None DOI: 10.4103/0028-3886.107927
Source of Support: This article is supported by the Youth Project No. 30801185 (to Y.L.) and the General Program No.30872673 (to C.Y.) from the National Natural Science Foundation of China, Conflict of Interest: None
Background: Recent genome-wide and locus-specific association studies identified RNF213 as an important Moyamoya disease (MMD) susceptibility gene. But the results of these studies are limited by the few subjects, different methodologies and ethnicities. Aims: To investigate the association between p.R4810K (rs 112735431, ss179362673; G > A) and p.R4859K (c.14576 G > A) polymorphisms of the RNF213 gene and MMD susceptibility. Settings and Design: We conducted a meta-analysis to evaluate the association. Materials and Methods: Two investigators independently searched the PubMed, Medline, and Embase databases for studies published before October 2012. For included studies, we performed meta-analyses using Cochrane RevMan software. Statistical Analysis: Summary odds ratios (ORs) and 95% confidence intervals (CIs) for RNF213 p.R4810K and p.R4859K polymorphisms; MMD were calculated in a fixed-effects model and a random effects model whenever appropriate. Results: Five eligible studies were reviewed and analyzed, which included two studies for p.R4810K polymorphisms (421 cases and 1214 controls) and three studies for p.R4859K polymorphisms (398 cases and 765 controls). Overall, the pooled results indicated that both p.R4810K polymorphisms and p.R4859K polymorphisms were associated with MMD risk (OR 92.03, 95% CI 54.06-156.65, P < 0.00001 and OR 157.53, 95% CI 85.37-290.7, P < 0.00001, respectively). Stratified analyses by ethnicity revealed the population attributable risks in the Japanese and Korean populations were larger than that in the Chinese population (P =0.0006). Conclusions: This meta-analysis demonstrated that there are strong associations between p.R4859K and p.R4810K polymorphisms of the RNF213 gene and MMD. The discoveries of its association with MMD may help in early diagnosis and prevention of this disease. Further study is still necessary to clarify the biochemical function and pathological role of RNF213 in MMD.
Keywords: Moyamoya disease, polymorphisms, RNF213
Moyamoya disease (MMD) is an uncommon cerebrovascular disorder that is characterized by progressive stenosis of the intracranial internal carotid arteries and their proximal branches. , Epidemiologically, MMD is more prevalent in east Asia than any region worldwide, which have been estimated at 0.94 per 100 000 in Japanese and 0.43 per 100,000 in Chinese. , Other factors such as a family history of the disease and gender are also associated with MMD.  These phenomena imply that genetic factors may play a potential important role in the pathogenesis of MMD. Although the incidence of MMD is not very high, it is an important cause of cerebral stroke in children and adults.  Generally, most children with MMD develop transient ischemic attack (TIA) or cerebral infarction, whereas about half of MMD adult patients develop intracranial bleeding and the other half develop TIA or cerebral infarction or both. In either case, MMD might lead to irreversible and devastating neurological deficits and intellectual impairments. ,, Therefore, prompt diagnosis and appropriate management are crucial to improve the long-term prognosis of patients, and identification of genetic variants associated with MMD may contribute to screening of susceptible MMD population.
Recently, several genome-wide and locus-specific association studies successfully identified a single important MMD susceptibility gene, ring finger protein 213 gene (RNF213). These studies indicated there was strongly association between RNF213 genotype, especially the RNF213 p.R4859K (c.14576 G > A) and p.R4810K (rs 112735431, ss179362673; G > A) polymorphism. [10-13] However, the results of these studies are limited by the few subjects, different methodologies, and ethnicities. Until recently, none systematic review or meta-analysis had been conducted to examine association between RNF213 p.R4859K and p.R4810K polymorphism and MMD risk. In order to derive a comprehensive estimation of the associations between RNF213 polymorphism and MMD risk, we conducted a meta-analysis to provide the best of evidence to assess the association between p.R4810K and p.R4859K polymorphisms of the RNF213 gene and MMD susceptibility.
Literature search strategy
We searched the PubMed, Medline (from 1966), and Embase (from 1980) for all articles on the association between RNF213 p.R4859K and p.R4810K polymorphism and MMD (last search update October 25, 2012) with the combination of the English key terms "ring finger protein 213" or "RNF213" or "p.R4859K" or "p.R4810K" or "polymorphism" or "genotype" or "allele," and "Moyamoya Disease" or "MMD." All published papers in English and Chinese language with available full text matching the eligible criteria were retrieved. The reference lists of all relevant papers and literature reviews were checked.
Study selection and inclusion criteria
Two review authors (JM and YL) independently screened the titles, abstracts, and keywords of citations obtained from the searches of the electronic databases and excluded studies that were clearly irrelevant. We obtained the full text of the remaining studies and the same two authors independently assessed which trials met the predefined inclusion criteria, including the following: 1) case-control studies were conducted to evaluate the association between RNF213 polymorphism and the risk of MMD; 2) sufficient genotype data were presented to calculate the odds ratios (ORs) and 95% confidence intervals (CIs); and 3) The paper should clearly describe the diagnosis of MMD and the sources of cases and controls. Major reasons for exclusion of studies were as follows: 1) duplicated studies; 2) review, editorial, or comment; and 3) limited sample size. Any disagreements were resolved by consensus between investigators.
The following data of included studies were extracted independently by two review authors (JM and CY): The first author's name, year of publication, country/region of participant, ethnicity, source of control group (population- or hospital-based controls), number of cases and controls, genotypes, genotyping methods, and evidence of Hardy-Weinberg equilibrium (HWE). Population-based case-control study was defined as control from healthy people, and hospital-based case-control study was defined as control from hospitalized patients. 
We performed meta-analyses using Cochrane RevMan (version 5.1) software.  The strength of the association between the risk of MMD and the RNF213 p.R4810K and p.R4859K polymorphisms were estimated using ORs, with the corresponding 95% CIs. For RNF213 p.R4810K and p.R4859K polymorphisms, we calculated directly study-specific ORs for the presence vs. absence of MMD, comparing participants with variant genotype (A/A or G/A) and wild-type G/G homozygote. If feasible, we also carried out the stratified analyses by ethnicity, country, and publication time. We assessed and quantified statistical heterogeneity for each pooled summary estimate using Cochran's Q statistic and the I 2 statistic, respectively. Substantial heterogeneity will be considered to exist with I >50% and Chi square test P < 0.1. Meta-analysis was performed via effect model (the Mantel-Haenszel method) if there was no evidence of statistical heterogeneity. The random-effects model was employed to pool studies when statistical heterogeneity occurred.
The combined search strategy identified 238 citations. After title, abstract, and full text screen, five case-control studies satisfied all inclusion criteria for final analysis, including two studies for p.R4810K and three studies for p.R4859K polymorphisms. ,,,, The literature search and study selection procedures are shown in [Figure 1].
The characteristics of selected studies are summarized in [Table 1]. Three studies were carried out in Japan and one study in China, including subjects of east Asian descent. One study was carried out in China, Japan, Korea, and Germany, including subjects of mixed descent. Diagnostic criteria of MMD in all included studies were based on the Japanese Research Committee on Moyamoya disease of the Ministry of Health, Welfare and Labour, Japan (RCMJ) criteria.  Controls were mainly from healthy populations and matched for age and/or sex. Controls in Miyawaki et al.  and 284 Caucasian controls in Liu et al.  were hospital-based. Most of studies extracted DNA from peripheral blood, and HRMA and Taqman PCR were mainly used for genotyping. The genotype distributions among the controls of all studies followed HWE except for one studies  for the RNF213 p.R4810K polymorphism.
Association between RNF213 p.R4810K polymorphism and MMD
As p.R4810K variant was not present among Caucasian cases or controls, we only calculated the OR of p.R4810K in the risk of MMD in east Asian. In the two studies with dichotomous data on MMD, pooled results with 421 cases and 1214 controls showed that, compared with people with a G/G genotype of RNF213 p.R4810K, those with an A+(A/G or A/A) genotype had significant increased odds of having MMD (OR 92.03, 95% CI 54.06-156.65; [Figure 2]). As there is substantial statistical heterogeneity: Chi square =1.94, df =1 (P =0.16); I =48%, we performed a stratified analysis according to the ethnicity (Japanese, Korean, and Chinese). The OR of A+ (A/G or A/A) genotype for MMD in Japanese was 338.94, 95% CI 150.32-764.20. The OR of A+ (A/G or A/A) genotype for MMD in Korean was 135.63, 95% CI 44.02-417.86. The OR of A+ (A/G or A/A) genotype for MMD in Chinese was 25.06, 95% CI 8.79-71.46 [Table 2]. Although p.R4810K was identified in three ethnic populations for significantly increasing the risk of MMD, the population attributable risks in the Japanese and Korean populations were larger than that in the Chinese population.
Association between RNF213 p.R4859K polymorphism and MMD
Three case-control studies with 398 cases and 765 controls for RNF213 p.R4859K were included eventually. [Table 2] listed the main results of this pooled analysis and [Figure 3] showed the association of MMD risk with RNF213 p.R4859K polymorphism in the form of forest plots. The results of combined analyses of all studies suggested that the RNF213 p.R4859K polymorphism was significantly associated with increased MMD risk (A/A or A/G vs. G/G, OR =157.53, 95% CI 85.37-290.70). There is not any heterogeneity: Chi square =2.04, df =2 (P =0.36); I =2%.
The present meta-analysis, including five case-control studies, explored the association between the p.R4859K and p.R4810K polymorphisms of the RNF213 gene and MMD risk. We found that there are strong association between p.R4859K and p.R4810K polymorphisms of the RNF213 gene and MMD (OR 92.03, 95% CI 54.06-156.65, P < 0.00001 and OR 157.53, 95% CI 85.37-290.70, P < 0.00001, respectively). These genetic variants associated with MMD not only provided an opportunity to screen of susceptible MMD population but also helped us to further investigate the pathogenesis of MMD.
It is interesting that the result of Liu et al. showed the p.R4810K variant was not present among Caucasian MMD cases or controls. The relatively high prevalence of p.R4810K among east Asians may account for the higher prevalence of MMD in east Asians than in Caucasians. In addition, p.R4810K was identified in three ethnic populations in east Asia, the population attributable risks in the Japanese and Korean populations were larger than that in the Chinese population. These pieces of evidence strongly suggest genetic heterogeneity (P =0.0006). Recent demographic surveys revealed that 1-2% of the Japanese and Korean general populations and about 0.4-0.5% of Chinese populations are carriers of these variants in the RNF213 gene.  Given that the prevalence of MMD in patients is about 1/10 carriers of these variants, only a minor portion of carriers will develop MMD. Thus, biochemical function and pathological role of RNF213 in MMD and any other susceptibility gene of MMD should be further investigated.
In the present meta-analysis, we searched as many publications as we could, and we believe that most of the related literature have been obtained and screened in our study. There is no evidence of significant heterogeneities in pooled analysis of each outcome. However, in this meta-analysis, four limitations should be noted: First, this paper was limited by analyzing two single nucleotide polymorphisms (SNPs), respectively, and lack of combination of two-SNP analysis. RNF213 enzyme activity might be affected by single or combination of p.R4859K and p.R4810K polymorphisms. Second, we must emphasize the fact that this meta-analysis adds more studies and increases the sample size compared with each single study, but the number of studies and the number of subjects in the meta-analysis were still small, and the increased sample size may not have been more important based on huge significance of OR. Third, we intended to access publication bias using funnel plot and linear regression test, however there were very few included studies to enable meaningful analysis. Fourth, by its restriction to fully published studies, our systematic review is potentially subject to publication bias.
There are strong associations between p.R4859K and p.R4810K polymorphisms of the RNF213 gene and MMD. The discoveries of its association with MMD may help for early diagnosis and prevention of MMD. Further study is still necessary to clarify the biochemical function and pathological role of RNF213 in MMD.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2]