Moyamoya disease is a progressive cerebrovascular disease that predominantly affects children and young adults.1 Although the pathogenesis of this disease is still not fully understood, a single nucleotide polymorphism in RNF213 (p.R4810K; c.14429G>A: rs112735431) was identified as a susceptibility factor in Asian countries, including Japan, in 2011.2,3 This point variant has also been identified in a substantial proportion of patients with underlying disease suggestive of quasi-moyamoya disease,4,5 and current treatment guidelines for moyamoya disease state that patients with underlying disease can be diagnosed with moyamoya disease in a broad sense.1 To date, many studies have identified the relationships of these genetic variants with clinical presentation and outcomes in patients with moyamoya disease.6–10 According to these previous studies, patients with homozygous variants of RNF213 p.R4810K are generally regarded to have more severe disease than patients with heterozygous variants or wild-type RNF213. However, most previous studies analyzed mixed groups of adult-onset and pediatric-onset patients, and to the best of our knowledge, few studies have focused on the relationship between genetic factors and clinical presentation, as well as prognosis, in pediatric patients alone. It is possible that the contributions of genetic and environmental factors differ between pediatric-onset and adult-onset patients, and separation of pediatric and adult patients may highlight the clinical significance of genetic factors in pediatric moyamoya disease.
In this single-center study, we focused on patients with pediatric-onset moyamoya disease and evaluated the relationships between the profile of the RNF213 p.R4810K variant and clinical presentation and prognosis. We also analyzed RNF213 variants other than p.R4810K in patients without the p.R4810K variant and focused on patients with very early onset occurring before 1 year of age (infantile onset).
Methods
This study was approved by the ethical committee of the local institutional review board, and written informed consent was obtained from all participants and/or their guardians after a detailed explanation of the study had been provided. At our institution, we have treated pediatric patients with moyamoya disease since 1979,11 and we started a genetic study in 2012.12–15 From March 2012 to August 2020, 129 patients with pediatric onset (≤ 15 years old), who initially presented to our department from 1979 to 2020, visited our hospital and agreed to participate in this study. All patients were Japanese and had been diagnosed with moyamoya disease on the basis of diagnostic criteria.1 An age cutoff of ≤ 15 years was used in this study because the Japanese Society of Pediatric Neurosurgery used this cutoff to define pediatric patients.16 Patients who had had previous episodes suggestive of transient ischemic attack (repeated ischemic symptoms, seizures, or involuntary movements) during childhood but were diagnosed with moyamoya disease after 15 years of age were regarded as adult-onset cases and excluded from the analysis.17 The analyzed patients consisted of 40.3% of the 320 patients who presented to our department from 1979 to 2020 and included several patients reported in our previous studies.12,13,15,18
Collection of Clinical Information
By retrospectively reviewing the prospective database of patients with moyamoya disease at our department, we collected the following information on each participant. Regarding initial presentation, we collected onset age, sex, clinical manifestation, initial MRA stage (highest stage on the right or left side),19 presence of posterior cerebral artery (PCA) lesions (on the right or left side), parenchymal lesions visible on initial MRI (cortical infarction, white matter infarction, white matter hyperintensity, and hemorrhage), family history of moyamoya disease, underlying disease suggestive of quasi-moyamoya disease,1 renal artery stenosis, systemic arterial stenosis other than renal artery stenosis, and initial modified Rankin Scale (mRS) score. At our institution, we screen for renal artery stenosis during every catheter angiography examination and as part of the workup for juvenile hypertension.
Regarding the clinical course, we recorded whether surgical treatment was performed, age at surgery, surgical outcome (see the Treatment Strategy of Our Institution section), and follow-up period. For patients with a follow-up period ≥ 1 year, cerebrovascular events that occurred during the follow-up period were also collected; these included novel symptomatic cerebral infarction (occurred in patients who did not initially present with infarction), novel symptomatic cerebral hemorrhage (occurred in patients who did not initially present with hemorrhage), de novo PCA lesions (occurred in patients who did not have PCA lesions on initial presentation), and mRS score at the last visit. Surgical outcome of the initial surgery was assessed in patients treated with multiple operations and defined as “good” if the ischemic symptoms from the surgically treated side disappeared and did not recur again for more than 1 year. Angiogenesis was confirmed with MRA20 or catheter angiography, and hemodynamic improvement was confirmed with perfusion MRI. As defined in a previous study,11 patients enrolled in ordinary school courses, workers in ordinary jobs, and housewives had mRS scores of 0–2; patients enrolled in special education classes or employed through employment services for handicapped people had an mRS score of 3; and physically disabled or deceased persons had mRS scores of 4–6, in accordance with the original definition (4, moderately severe physical disability and unable to walk; 5, severe disability and bedridden; 6, dead).
Treatment Strategy of Our Institution
At our department, we exclusively performed indirect revascularization surgery on pediatric and adult patients with moyamoya disease to mitigate hemodynamic impairment. The main surgical method is encephalo-duro-arterio-synangiosis (EDAS),21 but encephalo-duro-pericranio-synangiosis (EDPS) to the anterior or posterior region22 has also been selected for some cases, depending on the location of the ischemic area and the available donor artery.22 Until 2001, we performed xenon-enhanced CT and/or SPECT before and after surgery, and most patients were treated with bilateral EDAS if vascular lesions existed. Since 2001 when the attending neurosurgeon was changed at our institution, we have used perfusion MRI such as dynamic susceptibility contrast imaging23 and arterial spin labeling24 to decide on whom and where to perform indirect anastomosis (EDAS, EDPS, or both), depending on the era when the surgery took place. The surgical indication was not changed even if the patient had an underlying disease suggestive of quasi-moyamoya disease.1 Preoperative catheter angiography was always performed, but postoperative catheter angiography was rarely performed after 2001 because of its invasiveness and the increased availability of MRI.
In this study, we aimed to investigate the effect of indirect revascularization, and surgical outcome was regarded as good if ischemic symptoms from the surgically treated side disappeared and never recurred, angiogenesis was confirmed with MRA or catheter angiography, and hemodynamic improvement was confirmed with perfusion studies 1 year after surgery. Outcome was assessed for each patient by the attending neurosurgeon (T.N.) and the primary researcher (S.H.). For patients who underwent multiple operations, surgical outcome was assessed for the initial surgery alone. Perioperative infarction was not considered during the assessment of surgical outcomes in this study.
Identification of the RNF213 p.R4810K Variant
Genomic DNA was extracted from peripheral blood using Genomix 2.4 (Biologica) by the depository (H.U. Frontier). To identify the RNF213 p.R4810K variant, exon 60 of RNF213 was amplified with polymerase chain reaction (forward primer: CTCGCAGCCAGTCTCAAAGT; reverse primer: ATGTTTTTGGGGTTCAAGCA) using a PTC-100 Programmable Thermal Controller (MJR/Bio-Rad). Then, direct sequencing was performed using a BigDye Terminator v1.1 Cycle Sequencing Kit and an Applied Biosystems 3130xl Genetic Analyzer (Thermo Fisher Scientific) by another depository (Genome Laboratory, Tokyo Medical and Dental University). The genotype of each patient was categorized as homozygous mutant (A/A), heterozygous (G/A), or wild type (G/G). The investigators involved in genotyping had no prior knowledge of the patients’ clinical backgrounds.
Rare Variants and Their C-Scores in Patients Without the p.R4810K Variant
For patients without the p.R4810K variant, identification of rare variants other than p.R4810K was performed with Ion Torrent sequencing and/or whole-exome sequencing. The details of Ion Torrent sequencing and whole-exome sequencing were described in our previous reports.5,12 Briefly, Ion Torrent sequencing was performed with an Ion AmpliSeq primer pool (Thermo Fisher Scientific) that was custom-designed for all exons of the RNF213 gene. Whole-exome sequencing was performed with the SureSelect Human All Exon V5 kit (Agilent Technologies Inc.) and HiSeq 2000 sequencer (Illumina). The detected candidate variants were validated with standard polymerase chain reaction–based amplification, followed by BigDye terminator cycle sequencing on a 3130xl genetic analyzer (Thermo Fisher Scientific).
The functional effects of the detected variants (C-scores) were evaluated using Combined Annotation Dependent Depletion (CADD) version 1.3 (http://cadd.gs.washington.edu/).25
Statistical Analysis
Statistical analysis was performed using JMP version 12.0.1 (SAS Institute), and p < 0.05 was regarded as statistically significant. The Fisher’s exact test or analysis of variance was used to compare the three genotype groups, depending on the type of variable.
Results
Clinical Presentation and the RNF213 p.R4810K Variant
A summary of the results is presented in Table 1 and Fig. 1. More than 80% of the patients had either heterozygous or homozygous RNF213 p.R4810K variants. The onset age was significantly different among genotypes. Onset age was younger in the A/A and G/G groups compared with that of the G/A group, and the percentages of patients with onset age < 4 years was greater in these groups. Notably, all patients who developed moyamoya disease at less than 1 year of age had the G/G genotype for the RNF213 p.R4810K variant (Fig. 1, arrow).
Summary of pediatric patients with moyamoya disease evaluated in this study
| Characteristic | Genotype | p Value | |||
|---|---|---|---|---|---|
| A/A | G/A | G/G | All Patients | No Underlying Disease | |
| Patients | 12 (9.3) | 92 (71.3) | 25 (19.3) | ||
| Initial presentation | |||||
| Onset age, yrs | 4.2 ± 3.2 (1–10) | 7.1 ± 3.7 (1–15) | 4.4 ± 0.9 (0–15) | 0.034 | 0.017 |
| Onset at <4 yrs | 7 (58.3) | 16 (17.4) | 8 (32.0) | 0.008 | 0.003 |
| Female sex | 6 (50.0) | 59 (64.1) | 13 (52.0) | 0.407 | 0.564 |
| Clinical manifestation | |||||
| TIA | 7 (58.3) | 66 (71.7) | 9 (36.0) | 0.005 | 0.003 |
| Infarction | 3 (25.0) | 7 (7.6) | 6 (24.0) | 0.046 | 0.026 |
| Hemorrhage | 1 (8.3) | 2 (2.2) | 0 (0) | 0.310 | 0.343 |
| Epilepsy | 1 (8.3) | 7 (7.6) | 4 (16.0) | 0.481 | 0.435 |
| Other | 0 (0) | 10 (10.1) | 6 (24.0) | 0.052 | 0.068 |
| MRA stage | |||||
| 2 | 3 (25.0) | 14 (15.2) | 5 (20.0) | 0.397 | 0.476 |
| 3 | 8 (66.7) | 52 (56.5) | 16 (64.0) | ||
| 4 | 1 (8.3) | 26 (28.3) | 4 (16.0) | ||
| PCA lesion | 6 (50.0) | 34 (37.0) | 6 (24.0) | 0.267 | 0.282 |
| Parenchymal lesion | |||||
| Cortical infarction | 3 (25.0) | 25 (27.2) | 3 (12.0) | 0.248 | 0.405 |
| White matter infarction | 8 (66.7) | 28 (30.4) | 12 (48.0) | 0.026 | 0.027 |
| White matter hyperintensity | 7 (58.3) | 19 (20.7) | 10 (40.0) | 0.011 | 0.005 |
| Hemorrhage | 2 (16.7) | 5 (5.4) | 1 (4.0) | 0.383 | 0.424 |
| Family history of moyamoya disease | 4 (33.3) | 30 (32.6) | 5 (20.0) | 0.441 | 0.732 |
| Underlying disease | |||||
| Hashimoto disease | 0 (0) | 1 (1.1) | 0 (0) | 0.665 | |
| von Recklinghausen disease | 0 (0) | 1 (1.1) | 3 (12.0) | 0.043 | |
| Meningitis | 0 (0) | 0 (0) | 1 (4.0) | 0.191 | |
| Stenosis | |||||
| Renal artery | 0 (0) | 2 (2.2) | 1 (4.0) | 0.661 | 0.611 |
| Systemic arterial | 0 (0) | 0 (0) | 2 (8.0) | 0.035 | 0.027 |
| Initial mRS score | |||||
| 0–2 | 12 (100) | 87 (94.6) | 21 (84.0) | 0.108 | 0.124 |
| 3–5 | 0 (0) | 5 (5.4) | 4 (16.0) | ||
| Clinical course* | |||||
| Follow-up | |||||
| >1 yr | 11 (91.7) | 91 (98.9) | 21 (84.0) | 0.013 | 0.035 |
| Period, yrs | 12.5 ± 8.2 (1–26) | 12.2 ± 8.7 (1–34) | 10.6 ± 6.5 (1–23) | 0.708 | 0.718 |
| Revascularization surgery | 11 (100) | 77 (84.6) | 17 (81.0) | 0.147 | 0.119 |
| EDAS | 8 (72.7) | 62 (80.5) | 11 (64.7) | 0.347 | 0.346 |
| EDPS | 1 (9.1) | 2 (2.6) | 0 (0) | ||
| EDAS & EDPS | 2 (18.2) | 13 (16.9) | 6 (35.3) | ||
| Age at surgery | |||||
| All patients, yrs | 6.6 ± 4.8 (2–17) | 8.6 ± 4.7 (1–29) | 5.9 ± 2.9 (1–10) | 0.060 | 0.043 |
| ≤15 yrs | 10 (90.9) | 71 (92.2) | 17 (100) | 0.478 | 0.315 |
| Good surgical outcome | 10 (90.9) | 71 (92.2) | 13 (76.5) | 0.219 | 0.166 |
| De novo PCA lesion | 1 (9.1) | 8 (8.8) | 2 (9.5) | 0.902 | 0.117 |
| New infarction | 2 (18.2) | 7 (7.7) | 1 (4.8) | 0.897 | 0.522 |
| New hemorrhage | 1 (9.1) | 1 (1.1) | 1 (4.8) | 0.301 | 0.278 |
| Final mRS score | |||||
| 0–2 | 9 (81.8) | 87 (95.6) | 19 (90.5) | 0.200 | 0.365 |
| 3–5 | 1 (9.1) | 4 (4.4) | 2 (9.5) | ||
| 6 | 1 (9.1) | 0 (0) | 0 (0) | ||
TIA = transient ischemic attack.
Values are shown as number (percent) or mean ± standard deviation (range) unless indicated otherwise. Boldface type indicates statistical significance (p < 0.05).
Values for EDAS, EDPS, EDAS and EDPS, age at surgery, and good surgical outcome were calculated as percentages of patients who received revascularization surgery, and the values for revascularization surgery, de novo PCA lesion, new infarction, new hemorrhage, and final mRS score were calculated as percentages of patients who received follow-up for > 1 year.
Distribution of onset age for patients with each RNF213 p.R4810K genotype. All patients who presented with moyamoya disease at an age younger than 1 year (infantile onset, arrow) had the wild-type genotype (G/G). Figure is available in color online only.
Compared with the G/A group, the G/G group had an increased proportion of patients who presented with infarction and a reduced proportion who presented with transient ischemic attack. PCA lesions were more frequent in the A/A group and less frequent in the G/G group than in the G/A group, but the differences were not statistically significant. The proportions of patients with white matter infarction and hyperintensity on initial MRI were significantly different among the three groups, and the proportions of patients with these findings were increased in the A/A group and the G/G group in comparison with the G/A group. No significant association was observed between genotype and initial MRA stage, cortical infarction/hemorrhage on initial MRI, family history of moyamoya disease, or initial mRS score.
Underlying disease suggestive of quasi-moyamoya disease was more frequent in the G/G group than in the G/A group, and no patient in the A/A group had underlying disease. The significance of the difference among phenotypes remained unchanged when the analysis was limited to patients without underlying disease.
Renal artery stenosis was confirmed in the 1 patient with the G/G genotype and without underlying disease at the time of diagnosis and in the 2 patients with the G/A genotype and without underlying disease more than 10 years after initial surgery.15 Systemic arterial disease was observed in only 2 patients in the G/G group with infantile onset and without underlying disease.
Surgical and Clinical Outcomes and the RNF213 p.R4810K Variant
Of all 129 patients, 105 (81.0%) patients were surgically treated after initial presentation, and 123 (95.3%) patients, including all postoperative patients, were followed for ≥ 1 year. Age at surgery was significantly younger in the A/A and G/G groups after exclusion of patients with underlying disease (mean ± SD 6.6 ± 4.8 years in the A/A group, 8.7 ± 4.8 years in the G/A group, and 5.7 ± 3.0 years in the G/G group). Although all patients in the G/G group underwent surgery at age ≤ 15 years, good surgical outcomes were less frequently observed in the patients in the G/G group than in those in the G/A and A/A groups, but the difference was not statistically significant. Eleven patients (1 in the A/A group, 6 in the G/A group, and 4 in the G/G group) developed some degree of collateral vessels on MRA and hemodynamic improvement on perfusion MRI, but these patients had remnant transient ischemic symptoms 1 year after surgery. One patient in the G/G group with a history of meningitis developed almost no collateral vessels 1 year after surgery. Five patients (1 in the A/A group, 3 in the G/A group, and 1 in the G/G group) underwent additional surgery, and transient ischemic attacks disappeared in all but 1 patient. The remaining patients were conservatively treated because their transient ischemic attacks were not frequent or disappeared after antiplatelet treatment, or because the patient or the family requested conservative treatment.
The percentage of patients who were followed for ≥ 1 year was significantly greater in the G/A group than the A/A and the G/G group, but no significant difference in follow-up periods was observed after exclusion of patients who were followed for < 1 year. During the follow-up period, no significant correlation was observed between genotype and any cerebrovascular event. One patient in the A/A group died of brainstem infarction at the age of 39 years.18 Among the surviving patients at the time of analysis, fewer independent patients with final mRS scores ≤ 2 were in the A/A and G/G groups than in the A/G group, but the difference was not statistically significant.
Of the 6 patients with underlying disease, 2 patients in the G/A group (1 with Hashimoto disease and 1 with von Recklinghausen disease) and 3 patients in the G/G group (2 with von Recklinghausen disease and 1 with meningitis) were surgically treated, and all but 1 patient with a history of meningitis had good surgical outcomes. One patient (with von Recklinghausen disease) in the G/A group developed de novo PCA lesions and new infarctions during the follow-up period. All statistical significance regarding clinical outcome remained unchanged after exclusion of patients with underlying disease suggestive of quasi-moyamoya disease (p > 0.05).
Rare Variants and Their C-Scores in Patients Without the p.R4810K Variant
Eight other RNF213 variants were identified in 25 patients without the p.R4810K variant (Table 2). The CADD C-score of each variant, except 1, was greater than that of the p.R4810K variant (7.374); thus, these variants were regarded as having a more severe functional effect on this gene than p.R4810K. Two of these variants (p.H4058P and p.R4062Q) were reported in our previous study.13 Six of the 8 variants (other than p.A3468V and p.G3904V) were not listed in the gnomAD database (https://gnomad.broadinstitute.org/),26 the National Center for Biotechnology Information dbSNP147 database (http://www.ncbi.nlm.nih.gov/snp), or the Human Genetic Variation Database (http://www.hgvd.genome.med.kyoto-u.ac.jp/).27 Three patients without RNF213 variants carried the CCER2 gene variant, as reported in our previous study.12 None of the 4 patients with underlying disease (1 with meningitis and 3 with von Recklinghausen disease) had rare RNF213 variants.
Rare RNF213 variants other than p.R4810K identified in patients without the p.R4810K variant (G/G)
| Onset Age (yrs) | Sex | Clinical Manifestation & Clinical Course | MRA Stage | PCA Lesion | Parenchymal Lesions | Nucleotide Change (amino acid change) | CADD C-Score* |
|---|---|---|---|---|---|---|---|
| 0 | M | Epilepsy | 3 | No | CI, WMI, & WMH | c.12173A>C (p.H4058P)† | 29.90 |
| 0 | M | Infarction | 3 | No | CI, WMI, & WMH | c.12463A>C (p.T4155P) | 15.50 |
| 0 | F | Epilepsy | 3 | No | WMI | c.12394T>A (p.S4132T) | 29.70 |
| 2 | M | TIA | 3 | No | None | c.10403C>T (p.A3468V) | 6.109 |
| 7 | F | TIA | 3 | No | WMH | c.11711G>T (p.G3904V) | 17.51 |
| 8 | M | Headache | 3 | No | None | c.14248G>A (p.E4750K) | 23.50 |
| 11 | M | Asymptomatic to ICH | 3 | No | None | c.12185G>A (p.R4062Q)† | 31.00 |
| 13 | F | Asymptomatic | 3 | No | None | c.4876C>G (p.L1626V) | 24.10 |
CI = cortical infarction; ICH = intracerebral hemorrhage; WMH = white matter hyperintensity; WMI = white matter infarction.
All variants, except 1, had a higher CADD C-score than the p.R4810K variant (7.374).
Previously reported by Akagawa et al., 2018.13
Figure 2 summarizes the findings of 4 patients with infantile onset without the p.R4810K variant. Three of these 4 patients had heterozygous variants other than p.R4810K in the RNF213 gene. The C-scores of all 3 variants were greater than the C-score of p.R4810K (7.374) and were regarded as having a more severe functional effect on RNF213 than p.R4810K. These patients were referred to our department at various ages (7 months, 1 year 9 months, 1 year 5 months, and 9 years for patients 1–4, respectively) and had various degrees of cerebral infarction. Two of 4 patients had extracranial systemic arterial stenosis (the vertebral, thoracic and abdominal aorta, femoral, and pelvic arteries were affected in patient 2;13 the renal arteries and peripheral pulmonary arteries were affected in patient 3). All 4 patients were surgically treated (at the ages of 5, 5, 1, and 9 years for patients 1–4, respectively), had good surgical outcomes, and had no subsequent cerebrovascular events for more than 7 years (10, 13, 7, and 21 years for patients 1–4, respectively). However, 3 of 4 patients (75%) had persistent motor impairment and cognitive decline, and only 1 of 4 patients (patient 4) had an independent life with a final mRS score ≤ 2.
Details of the 4 patients who presented with moyamoya disease at an age younger than 1 year (infantile onset). Initial images obtained with MRA, T2-weighted imaging (T2WI), or fluid-attenuated inversion recovery (FLAIR) are presented, as well as information regarding initial presentation. Three of these 4 patients had rare variants with higher C-scores than the p.R4810K variant (7.374). CI = cortical infarction; WMH = white matter hyperintensity; WMI = white matter infarction. †Previously reported by Akagawa et al., 2018.13
Discussion
This single-center study of a relatively large volume of Japanese pediatric patients with moyamoya disease revealed that patients with negative results for the RNF213 p.R4810K variant were more likely to present with infarction and less likely to present with transient ischemic attack, which is the most frequent presentation in this disease population,1 than patients with heterozygous variants. Eight of 25 patients without the p.R4810K variant had variants other than RNF213 p.R4810K. Notably, all 4 patients with infantile onset lacked the p.R4810K variant, and 3 of 4 had RNF213 variants with more severe functional effects than the p.R4810K variant.
No previous studies have reported that patients with wild-type RNF213 p.R4810K (i.e., G/G genotype) have a more severe form of moyamoya disease than heterozygous (G/A) patients, as in our study. Homozygous (A/A) patients reportedly have younger onset age and greater percentages of homozygous patients present with infarction and PCA lesions according to studies from Japan,6 Korea,10 and China,8,9,28 which are grossly in accord with our results. However, these studies reported better clinical presentation in G/G patients than G/A patients, which was the opposite of our results. The difference may have resulted from the mixture of patients with adult-onset and pediatric-onset disease in previous studies, whereas we focused on only patients with pediatric onset. The prevalence of pediatric onset in G/G patients was reportedly lower than that in G/A patients,9 so the mixture of pediatric-onset and adult-onset patients may have attenuated the true nature of pediatric-onset disease in patients with the G/G genotype. By focusing on patients with pediatric onset, we found that our G/G group, as well as the A/A group, had younger onset age and increased proportions of patients with infarction as initial presentation and white matter infarction and hyperintensity on the initial MRI in comparison with G/A group. Separating pediatric-onset and adult-onset patients may be necessary to fully understand the contributions of genetic variants to pediatric moyamoya disease.
The findings of our patients with infantile onset suggest that genetic factors other than the RNF213 p.R4810K variant may induce a more severe form of pediatric moyamoya disease than that seen in patients with the heterozygous p.R4810K variant. The absence of p.R4810K variants suggests that two possible factors induced moyamoya disease: either environmental factors or genetic variants other than the RNF213 p.R4810K variant. Family history of moyamoya disease, which strongly suggests genetic factors rather than environmental factors, was not significantly less common among G/G patients than A/A and G/A patients. This result conflicts with previous studies that reported a lower prevalence of family history in the G/G group than the A/A and A/G groups,6,8,10 but again, the difference may be due to the mixture of adult- and pediatric-onset patients in previous studies. Three of 4 patients with infantile onset had rare RNF213 variants with more severe functional effects than p.R4810K. These results suggest that the more pathological functional effect of rare variants may be, at least in some part, the cause of the more severe clinical course of pediatric moyamoya disease. However, the functional effects of these genetic variants seem insufficient to explain severity of disease, because not all patients with rare variants with high CADD C-scores had severe clinical manifestations (Table 2). Some environmental or genetic factors other than RNF213 may be related to the onset of pediatric moyamoya disease, too.
We did not find any significant relationship between genotype of the p.R4810K variant and cerebrovascular events during the follow-up period, although the occurrence of cerebrovascular events was comparable to those reported by previous studies.29,30 The small number of patients and the total number of cerebrovascular events may partly explain these results, or these results may suggest that the p.R4810K genotype alone is insufficient to predict clinical outcomes in patients with pediatric moyamoya disease. Previous studies from Japan and China reported that the p.R4810K genotype is not a predictor of long-term outcomes, including cerebrovascular events during follow-up and final mRS score,28,31 but the study from China suggested a lower recurrent stroke rate in G/A patients than G/G and A/A patients.28 As discussed above, rare variants other than p.R4810K and other genetic variants may also be related to long-term outcome as well as initial presentation, and the p.R4810K genotype alone may not be a predictor of long-term outcome.
Although not statistically significant, good surgical outcome after indirect revascularization surgery was less frequently observed in G/G patients than G/A and A/A patients. The outcome of indirect revascularization is considered reliant on patient-specific ability for collateral recruitment,29 and a study from China found that better indirect anastomosis developed after revascularization surgery in G/A patients than G/G patients.32 It is possible that the RNF213 p.R4810K variant has a protective role in patients with moyamoya disease by enhancing the “conversion of the internal carotid system to the external carotid system”32 and producing better postoperative anastomosis. Statistical significance may not have been obtained in our study because of the relatively small number of patients who underwent revascularization surgery. Because postoperative catheter angiography was not performed on most patients, we could not quantitatively assess indirect revascularization; therefore, in the future we plan to perform a quantitative assessment of recent imaging studies (MRA and arterial spin labeling) to assess the effects of different genotypes on indirect anastomosis. Inclusion of more participants would further clarify the possible role of the RNF213 p.R4810K variant in prediction of surgical outcome after indirect revascularization surgery.
The limitations of this study include its single-center design that only included a small portion of the pediatric patients with moyamoya disease in Japan. Less than half the patients who presented to our department during the study period were included. We performed a detailed analysis of only RNF213 in G/G patients, but even patients with the G/A phenotype may have had compound heterozygous variants that may have affected clinical presentation and outcome.33 However, the reported proportion of patients with compound heterozygous variants is low, and the lack of information regarding compound heterozygous variants did not affect the severe pathological implications of the G/G phenotype in this study. Genetic variants other than RNF213 may also have affected clinical presentation and outcome,12,34 but we did not evaluate these variants in every patient. Because of the different prevalence rates of genetic variants among countries,8–10 the results cannot be directly applied to different countries and need validation in each country. Nevertheless, this is currently the only study to show that a negative result for the RNF213 p.R4810K variant may be a novel biomarker of the severe form of pediatric moyamoya disease. A multicenter study with a large volume of patients throughout the country35 is warranted to determine the role of genetic analysis in the treatment of Japanese patients with pediatric moyamoya disease.
Conclusions
By focusing on pediatric-onset patients, we found that the absence of the RNF213 p.R4810K variant may be a novel biomarker for identification of patients with severe clinical presentation of pediatric moyamoya disease.
Acknowledgments
We thank Yaeko Furuhashi and Yuichi Yatoh for helping us identify the RNF213 p.R4810K variants. This work was supported by KAKENHI grants from the Japan Society for the Promotion of Science (nos. 23592109 and 19K09537).
Disclosures
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.
Author Contributions
Conception and design: Hara, Mukawa, Nariai. Acquisition of data: Hara, Mukawa, Akagawa, Thamamongood, Inaji, Tanaka, Nariai. Analysis and interpretation of data: Hara, Mukawa, Nariai. Drafting the article: Hara. Critically revising the article: Mukawa, Akagawa, Thamamongood, Inaji, Tanaka, Maehara, Kasuya, Nariai. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Hara. Statistical analysis: Hara. Administrative/technical/material support: Mukawa, Akagawa, Kasuya, Nariai. Study supervision: Maehara, Kasuya, Nariai.
Supplemental Information
Previous Presentations
The results of this study were presented at the 49th Annual Meeting of The Japanese Society of Pediatric Neurosurgery, Fukushima, Japan, July 5, 2021.
References
- 1↑
Tominaga T, Suzuki N, Miyamoto S, Koizumi A, Kuroda S, Takahashi JC, et al. Recommendations for the Management of Moyamoya Disease: A Statement from Research Committee on Spontaneous Occlusion of the Circle of Willis (Moyamoya Disease). [2nd Edition]. Surg Cereb Stroke. 2018;46(1):1–24.
- 2↑
Liu W, Morito D, Takashima S, Mineharu Y, Kobayashi H, Hitomi T, et al. Identification of RNF213 as a susceptibility gene for moyamoya disease and its possible role in vascular development. PLoS One. 2011;6(7):e22542.
- 3↑
Kamada F, Aoki Y, Narisawa A, Abe Y, Komatsuzaki S, Kikuchi A, et al. A genome-wide association study identifies RNF213 as the first moyamoya disease gene. J Hum Genet. 2011;56(1):34–40.
- 4↑
Morimoto T, Mineharu Y, Kobayashi H, Harada KH, Funaki T, Takagi Y, et al. Significant association of the RNF213 p.R4810K polymorphism with quasi-moyamoya disease. J Stroke Cerebrovasc Dis. 2016;25(11):2632–2636.
- 5↑
Nomura S, Akagawa H, Yamaguchi K, Ishikawa T, Kawashima A, Kasuya H, et al. Rare and low-frequency variants in RNF213 confer susceptibility to moyamoya syndrome associated with hyperthyroidism. World Neurosurg. 2019;127:e460–e466.
- 6↑
Miyatake S, Miyake N, Touho H, Nishimura-Tadaki A, Kondo Y, Okada I, et al. Homozygous c.14576G>A variant of RNF213 predicts early-onset and severe form of moyamoya disease. Neurology. 2012;78(11):803–810.
- 7
Zhang Q, Ge P, Ma Y, Zhang D, Wang R, Zhang Y, et al. Clinical features and surgical outcomes of patients with moyamoya disease and the homozygous RNF213 p.R4810K variant. J Child Neurol. 2019;34(13):793–800.
- 8↑
Zhang Q, Liu Y, Zhang D, Wang R, Zhang Y, Wang S, et al. RNF213 as the major susceptibility gene for Chinese patients with moyamoya disease and its clinical relevance. J Neurosurg. 2017;126(4):1106–1113.
- 9↑
Wang Y, Zhang Z, Wei L, Zhang Q, Zou Z, Yang L, et al. Predictive role of heterozygous p.R4810K of RNF213 in the phenotype of Chinese moyamoya disease. Neurology. 2020;94(7):e678–e686.
- 10↑
Kim EH, Yum MS, Ra YS, Park JB, Ahn JS, Kim GH, et al. Importance of RNF213 polymorphism on clinical features and long-term outcome in moyamoya disease. J Neurosurg. 2016;124(5):1221–1227.
- 11↑
Mukawa M, Nariai T, Matsushima Y, Tanaka Y, Inaji M, Maehara T, et al. Long-term follow-up of surgically treated juvenile patients with moyamoya disease. J Neurosurg Pediatr. 2012;10(5):451–456.
- 12↑
Mukawa M, Nariai T, Onda H, Yoneyama T, Aihara Y, Hirota K, et al. Exome sequencing identified CCER2 as a novel candidate gene for moyamoya disease. J Stroke Cerebrovasc Dis. 2017;26(1):150–161.
- 13↑
Akagawa H, Mukawa M, Nariai T, Nomura S, Aihara Y, Onda H, et al. Novel and recurrent RNF213 variants in Japanese pediatric patients with moyamoya disease. Hum Genome Var. 2018;5:17060.
- 14
Aoyama J, Nariai T, Mukawa M, Inaji M, Tanaka Y, Maehara T. Case of familial moyamoya disease presenting 10 years after initial negative magnetic resonance screening in childhood. World Neurosurg. 2017;105:1035.e1–1035.e4.
- 15↑
Hara S, Shimizu K, Nariai T, Kishino M, Kudo T, Umemoto T, et al. De novo renal artery stenosis developed in initially normal renal arteries during the long-term follow-up of patients with moyamoya disease. J Stroke Cerebrovasc Dis. 2020;29(8):104786.
- 16↑
Kanamori F, Araki Y, Yokoyama K, Uda K, Nishihori M, Izumi T, et al. Brain compression by encephalo-myo-synangiosis is a risk factor for transient neurological deficits after surgical revascularization in pediatric patients with moyamoya disease. World Neurosurg. 2020;133:e558–e566.
- 17↑
Yamamoto S, Kashiwazaki D, Uchino H, Saito H, Akioka N, Kuwayama N, Kuroda S. Clinical and radiological features of childhood onset adult moyamoya disease: implication for hemorrhagic stroke. Neurol Med Chir (Tokyo). 2020;60(7):360–367.
- 18↑
Mukawa M, Nariai T, Inaji M, Tamada N, Maehara T, Matsushima Y, et al. First autopsy analysis of a neovascularized arterial network induced by indirect bypass surgery for moyamoya disease: case report. J Neurosurg. 2016;124(5):1211–1214.
- 19↑
Houkin K, Nakayama N, Kuroda S, Nonaka T, Shonai T, Yoshimoto T. Novel magnetic resonance angiography stage grading for moyamoya disease. Cerebrovasc Dis. 2005;20(5):347–354.
- 20↑
Houkin K, Nakayama N, Kuroda S, Ishikawa T, Nonaka T. How does angiogenesis develop in pediatric moyamoya disease after surgery? A prospective study with MR angiography. Childs Nerv Syst. 2004;20(10):734–741.
- 21↑
Matsushima Y, Inaba Y. Moyamoya disease in children and its surgical treatment. Introduction of a new surgical procedure and its follow-up angiograms. Childs Brain. 1984;11(3):155–170.
- 22↑
Nariai T. Indirect bypass surgery for moyamoya disease. In: Kuroda S, ed. Moyamoya Disease: Current Knowledge and Future Perspectives. Springer Nature;2021:301–310.
- 23↑
Hirai S, Inaji M, Tanaka Y, Hara S, Nariai T, Maehara T. Correlation between clinical presentations and hemodynamic parameters measured by dynamic susceptibility contrast magnetic resonance imaging in adult patients with moyamoya disease. J Stroke Cerebrovasc Dis. 2017;26(12):2814–2820.
- 24↑
Hara S, Tanaka Y, Ueda Y, Hayashi S, Inaji M, Ishiwata K, et al. Noninvasive evaluation of CBF and perfusion delay of moyamoya disease using arterial spin-labeling MRI with multiple postlabeling delays: comparison with 15O-gas PET and DSC-MRI. AJNR Am J Neuroradiol. 2017;38(4):696–702.
- 25↑
Kircher M, Witten DM, Jain P, O’Roak BJ, Cooper GM, Shendure J. A general framework for estimating the relative pathogenicity of human genetic variants. Nat Genet. 2014;46(3):310–315.
- 26↑
Karczewski KJ, Francioli LC, Tiao G, Cummings BB, Alföldi J, Wang Q, et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature. 2020;581(7809):434–443.
- 27↑
Higasa K, Miyake N, Yoshimura J, Okamura K, Niihori T, Saitsu H, et al. Human genetic variation database, a reference database of genetic variations in the Japanese population. J Hum Genet. 2016;61(6):547–553.
- 28↑
Ge P, Ye X, Liu X, Deng X, Wang R, Zhang Y, et al. Association between p.R4810K variant and long-term clinical outcome in patients with moyamoya disease. Front Neurol. 2019;10:662.
- 29↑
Ravindran K, Wellons JC, Dewan MC. Surgical outcomes for pediatric moyamoya: a systematic review and meta-analysis. J Neurosurg Pediatr. 2019;24(6):663–672.
- 30↑
Kimiwada T, Hayashi T, Shirane R, Tominaga T. Posterior cerebral artery stenosis and posterior circulation revascularization surgery in pediatric patients with moyamoya disease. J Neurosurg Pediatr. 2018;21(6):632–638.
- 31↑
Nomura S, Yamaguchi K, Akagawa H, Kawashima A, Moteki Y, Ishikawa T, et al. Genotype-phenotype correlation in long-term cohort of Japanese patients with moyamoya disease. Cerebrovasc Dis. 2019;47(3-4):105–111.
- 32↑
Ge P, Zhang Q, Ye X, Liu X, Deng X, Wang J, et al. Angiographic characteristics in moyamoya disease with the p.R4810K variant: a propensity score-matched analysis. Eur J Neurol. 2020;27(5):856–863.
- 33↑
Moteki Y, Onda H, Kasuya H, Yoneyama T, Okada Y, Hirota K, et al. Systematic validation of RNF213 coding variants in Japanese patients with moyamoya disease. J Am Heart Assoc. 2015;4(5):e001862.
- 34↑
Wang X, Wang Y, Nie F, Li Q, Zhang K, Liu M, et al. Association of genetic variants with moyamoya disease in 13 000 individuals: a meta-analysis. Stroke. 2020;51(6):1647–1655.
- 35↑
Kazumata K, Ito M, Uchino H, Nishihara H, Houkin K. Proposal for a prospective registry for moyamoya disease in Japan. Neurol Med Chir (Tokyo). 2017;57(2):66–72.
