Meta-analysis of the association between RNF213 polymorphisms and clinical features of moyamoya disease in Asian population

.


Introduction
Moyamoya disease (MMD), also known as spontaneous occlusion of the Circle of Willis, is a chronic progressive cerebrovascular stenosis or occlusion disease.This situation is more prevalent in individuals from East Asian countries than those from the western countries, and can lead to severe loss of neurological function and cognitive disorders [1].
Although many progress have been made in our understanding of MMD, the etiology is still not well uncovered.The variant p.R4810K of RNF213 (corresponding to c .14429G > A p.R4810K; GenBank accession number NG_031980), was first identified as a pathogenic gene in the Japanese, Chinese and Korean patients in 2011, followed by more studies further resonating the result [2][3][4][5].
We proposed to collect the information about the RNF213 variants in non-Asian population as many as possible, but the frequency of the other variants was much lower and more diverse in non-Asian descent.So we only focused on the Asian population in this study.
Several studies have shown an association between RNF213 variants and susceptibility to MMD [4,5].The relationship between single locus mutation of RNF213 and clinical manifestations of MMD has also been reported [10].However, very few studies are available to elaborate the association between clinical features of MMD and SNPs of RNF213, so we conducted this meta-analysis to explore their correlation and provide evidence for clinical practice.

Literature retrieval strategy
Databases including PubMed, Google Scholars, EMBASE, Scopus and Cochrane Library were conducted from inception to May 15th 2022.We systematically reviewed the studies of RNF213 polymorphisms in MMD.The keywords combinations of Medical Subject Headings (MeSH) terms with the known gene name was("RNF213"[tw] OR "ring finger protein 213"[tw]) and ("Moyamoya Disease"[MeSH] OR "Moyamoya[tw]" OR "primary moyamoya"[tw] OR "primary moyamoya disease*"[tw]) and ("polymorphism, genetic"[MeSH] OR "Gene* Polymorphism*"[tw] OR "Polymorphism* (Genetics)"[tw] OR "Point Mutation"[MeSH] OR ("point mutation*"[tw] OR "mutation* point"[tw]).The term "variation" is increasingly used in human genetics, but we did not include this term based on past experience when constructing search formulas.To make the electronic search relevant and perfect, we also manually searched references listed in selected articles to identify other potentially eligible studies.The search was limited to the Asian region but there are no language restrictions.

Selection criteria and identification of relevant studies
We included related studies to evaluate associations between the RNF213 polymorphisms and clinical features of MMD.Publications that satisfied the criteria were included in a meta-analysis following the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement, which was published previously [11].The PRISMA flow diagram was used to determine whether to include or exclude any publication found in the literature according to the selected keywords (Fig. 1).
Correspondingly, the selected studies in our meta-analysis must meet the following criteria: Labour Sciences Research Grant for Research on Measures for Intractable Diseases in Japan [12], including unilateral MMD and definite MMD; (3) Data on clinical phenotypic characteristics in every study corresponding to specific genotypes must be clearly shown or indirectly calculated; (4) All variants incorporated in the meta-analysis should be explored in at least two published studies.For duplicate or overlapped publications existed, we chose the one with largest sample or the latest study.Studies that did not meet any of the above inclusion criteria were excluded.

Data extraction
Data were extracted from the selected studies by two authors independently and differences were resolved by discussion or a third researcher.For each included study, the following information was extracted: first author, publication year, study population (nation), the number of clinical phenotypes corresponding to the mutant type that including heterozygous and homozygous mutations and the wild type in each mutation site of RNF213 (Supplementary Tables 1-9).
The wild type of RNF213 is BB, the heterozygous mutant is AB and the homozygous mutant is AA.Therefore, we used published information on homozygous mutation, heterozygous mutation and wild types of the RNF213 polymorphisms in MMD to estimate the number of cases using a three model (AA+ BA vs BB, AB vs BB, AA vs BB).Three genetic models were compared to explore the relationship between RNF213 polymorphisms and clinical features of MMD.
We evaluated the quality of non-randomized studies using the Newcastle-Ottawa scale (NOS) [13].A study can be awarded from one to nine scores, the scores of 1-3,4-6, and 7-9 indicate low-, intermediate, and high-quality studies, respectively.Studies scoring ≥7 of the NOS were considered to be of high quality.

Data synthesis and publication bias
STATA 16.0 MP were applied to finish the statistical analysis.Binary variables were expressed by odds ratio, and continuous variables were expressed by mean ± standard difference.ORs and 95 % CIs were used to evaluate the association between RNF213 variants and clinical characteristics of MMD patients based on the clinical phenotype in mutant type and wild type.If I 2 < 50 % and P > 0.01, a fixed effect model would be adopted, otherwise a random-effect model would be performed.If I 2 > 75 %, then a sensitive analysis or subgroup analysis or meta-regression would be performed to identify the source of heterogeneity.
Considering that comparison model still showed high heterogeneity after subgroup analysis, we continued to perform sensitivity analysis to try to find the source of heterogeneity and assess the stability of the results by precluding individual studies one by one.Egger's linear regression test were used to estimate the publication bias in the studies included (P > 0.05 was considered to have no statistically significant publication bias).

Standard protocol approvals, registrations, and patient consents
There were no direct human participants that required approval by an institutional ethics committee in the study, for resembling reasons, no written informed consent was required.

Characteristics of the included studies
A total of 16 articles (3061 patients) were included in this metaanalysis to explore the association between RNF213 variants and clinical features of MMD [3,6,8,9,[14][15][16][17][18][19][20][21][22][23][24][25][26].All NOS scales included in the study had a high score (Table 1).The sample size of each independent study ranged from 20 to 1385.In this meta-analysis, the interest was given to five RNF213 rare variants (rs112735431, rs760732823, rs371441113, rs148731719 and rs9916351) and 9 various clinical features of MMD (Table 2).The populations studied were all Asian, including Chinese, Japanese, Korean and Indian.The rs112735431 polymorphism was common in four countries, and the variant of rs148731719 were found both in South Korea and in China.The two SNPs of rs760732823 and rs371441113 were only appeared in South Korea, another polymorphism of rs9916351 was only reported in China (Fig. 2).

Association of RNF213 polymorphisms with gender distribution
A total of 11 studies, two RNF213 missense variants (rs112735431 and rs760732823) were included to assess the effect of RNF213 polymorphisms on gender distribution in MMD.The results suggested no differences on gender distribution associated with RNF213 polymorphisms in MMD were observed (Table 3, Supplementary figure 1).Subgroup analysis on each polymorphism found the same result as well (Supplementary figure 1).

Association between RNF213 polymorphism and clinical features of MMD
As shown in Table 3, compared with wild type, the patients under 18 years of age at onset, familial MMD (AA+AB vs BB, AB vs BB and AA vs BB: P = 0.000，respectively), cerebral ischemic stroke (AA+AB vs BB and AB vs BB: P = 0.000，respectively) and posterior cerebral artery involvement (PCi) (AA+AB vs BB and AB vs BB: P = 0.000, respectively; AA vs BB: P = 0.006) were significantly more common in mutant type of RNF213.

RNF213 polymorphisms associated with cerebral stroke
A total of 13 selected studies, including two SNPs (rs112735431 and rs148731719) were analyzed to assess the association between RNF213 polymorphisms and ICH/IVH in MMD.The results indicated that no association between RNF213 variants and ICH/IVH of MMD (Table 3).There was significant heterogeneity in the included studies excepting in group of AA vs BB (Table 3).

Association between RNF213 polymorphisms and angiographic findings
A total of ten studies, involving two SNPs rs112735431and rs148731719, were included to evaluate the association of RNF213 polymorphisms and involvement of the posterior cerebral artery (PCA) in MMD in this meta-analysis.The results clearly showed that the mutation in RNF213 was associated with PCi in MMD with three comparative mutation models (AA+AB vs BB: OR=1.67, 95 % CI=1.40-2.00,AB vs BB: OR=1.65, 95 % CI=1.38-1.97;AA vs BB: OR=2.79, 95 % CI=1.35-5.79;Table 3).

Publication bias
From the Egger's test, we found no obvious publication bias in each outcome.(Supplementary table 10).

Discussion
As far as we know, this study was the first meta-analysis to evaluate RNF213 variants and clinical manifestations of MMD.The main results confirmed that RNF213 variants increased the risk of initial symptoms of cerebral ischemia and PCi of MMD in Southeast Asia.Besides, more MMD cases with other valuable clinical characteristic, such as early onset (< 18 years), family history of the disease were more common in the RNF213 variants.Subgroup analysis of different sites of RNF213 showed that the rs112735431 might decrease the risk of ICH/IVH, while the rs148731719 might was the opposite.
These results support the original hypothesis that RNF213 variants might be a valid bio-marker to predict the different clinical features of MMD.In addition, although some clinical homogeneity have been addressed in previous and present studies, this study also found the heterogeneity of clinical characteristics caused by SNPs of RNF213.
Therefore, RNF213 variants might cause significant differences in predicting the clinical feature of patients with MMD in Asia.The gender distribution in RNF213 variants was not significantly different, which was consistent with the findings of the previous study [3,10,18].Subsequent subgroup analyses based on gene variants showed no evidence of association on gender distribution neither in rs112735431 nor in rs148731719 (Table 3).
Several previous researches showed that the variants of rs112735431 could lead to significantly earlier-onset cases (age at onset < 18years) of MMD in patients compared to wild type, which has been further confirmed in the present analysis, especially at two SNPs rs112735431 and rs991635 [8,17,18,26,27].Interestingly, this SNPs of rs371441113 in Korea displayed the age of late-onset in MMD.
Nevertheless, heterozygous variant in rs112735431 was also associated with relatively late onset age among siblings of the Japanese family [28].However, the results indicated the variant of RNF213 could lead to the early onset of MMD.Furthermore, the results of sensitivity analysis did not change the conclusion, showing that our results were robust.Meanwhile, we also performed a subgroup analysis suggesting that rs112735431 was still associated with early onset, consequently we deduced that this could be due to the following reasons: (1) genetic heterogeneity; (2) differences between demographic and ethnic in included study; (3) environmental factors.
Family history is an important factor to the onset of MMD, which presents the characteristics of family aggregation in East Asia [29].Our study showed a statistically significant correlation between RNF213 variants and increased risk of familial MMD, which was in congruity with some previous studies [3].Due to the presence of mild heterogeneity, subgroup analysis and sensitivity analysis were conducted.Subgroup analysis revealed that the heterogeneity came from the site of rs112735431, which was considered to be related to differences in population and region.However, the results of sensitivity analysis still strongly supported that those patients with rs112735431 had higher risk of familial occurrence of the disease.As a consequence, we believe that RNF213 screening should be executed to identify asymptomatic patients with a hypothetical population showing high penetrance, particularly those who had a family history of MMD in Asian descent.
Pooled analysis in this study displayed that RNF213 variants could increase the risk of PCi in patients with MMD.Sensitivity analysis by successively excluding included studies found that the conclusions were stable.Consistent with the results from previously researchers, the rs112735431 polymorphism could increase the risk of PCi of MMD [15,17,18,26].They found that the patients with PCi at initial onset had significantly worse outcomes since this study has shown that steno-occlusive PCA lesions had significant impact on the severity of ischemia [30].Nevertheless, they have demonstrated that irreversible cerebral infarction was associated with greater risk of adverse clinical prognosis [31].Therefore, we speculated that the involvement of PCi could not provide effective compensation for cerebral blood circulation, thus leading to severe cerebral ischemia events, which still needs to be further verified.

Limitations
First, RNF213 is a crucial important susceptible gene to MMD.In this study, only populations in Asia were included, while populations in Europe and other regions were not.The participants included in this analysis may not be representative of the universal population.A recent study showed that RNF213 variants are associated with younger age of onset in non Asian MMD patients [27].Secondly, the exome Variant Server database reports 397 variants in RNF213, only 5 variants were included in this study, which were far from representative of the whole mutation characteristics of RNF213.Whereas the rs112735431 polymorphism included more studies, which could lead to site stratification.Thirdly, our subgroup analysis and sensitivities analysis cannot completely solve the source of inter-study heterogeneity.Therefore, more SNPs data need to be included to explain the possible causes of heterogeneity.Finally, the number of included cases for headache and seizure were too small, future studies need to increase the sample size for further discussion.

Conclusions
This meta-analysis provided evidence that RNF213 variants not only increase the risk of initial symptoms of cerebral ischemia and PCi of MMD in Southeast Asia, but also have early onset and a family history of complications.More attention should be paid to patients who have ischemic MMD occurred younger than 18 years old.RNF213 variant screening and cerebrovascular imaging examination should be performed to evaluate intracranial vascular involvement, so as to achieve early detection and early treatment and avoid more serious cerebrovascular event.

Fig. 1 .
Fig. 1.PRISMA flowchart describing the process of selecting the study.

Fig. 2 .
Fig. 2. The population distribution of various variants of RNF213 in multipleAsian countries selected for the study.

Table 1
Study quality assessment using the Newcastle-Ottawa tool.

Table 2
Correlation between five sites of RNF 213 polymorphisms and clinical characteristics of MMD.

Table 3
Association of clinical features with the genotype of RNF213 polymorphisms in Asian patients with MMD.