Moyamoya Disease and Monogenic Moyamoya Angiopathy

Progressive steno-occlusive arteriopathy of the terminal portion of the internal carotid arteries, with a basal collateral network of fragile vessels, which arises as an idiopathic disease with genetic susceptibility (RNF213), as a Mendelian disorder (ACTA2, GUCY1A1, BRCC3, ANO1, NOS3 and other genes) or as moyamoya syndrome associated with systemic genetic diseases.

Literature and regulatory information updated: 31 October 2026.

0. Initial Reference Table

1. Introduction

1.1. Original description and name

The angiographic changes that would come to define the disease were described in Japan in 1957 as bilateral "hypogenesis" of the internal carotid arteries, and the entity became established during the 1960s [40]. In 1969, Jiro Suzuki and Akira Takaku published, in English, the series that gave it the name moyamoya disease, from the angiographic appearance of the network of collateral vessels at the base of the brain, resembling a "puff of smoke" ("moyamoya", in Japanese, something hazy like cigarette smoke drifting in the air) [1, 6, 18]. The angiographic description brings together three elements: progressive stenosis or occlusion of the terminal portion of the internal carotid arteries (ICA) and the proximal portions of the anterior (ACA) and middle (MCA) cerebral arteries; a network of small dilated collateral vessels (lenticulostriate and thalamic perforators) near the lesions; and, as the occlusion progresses, regression of these vessels and transfer of hemispheric perfusion to the external carotid and vertebrobasilar systems [2, 40].

In 1974 the Research Committee on Moyamoya Disease (Research Committee on Moyamoya Disease [Spontaneous Occlusion of the circle of Willis]) was established in Japan; it set the first diagnostic criteria in 1978 and revised them in 1987, 1995, 2009, 2015 and 2021 [1]. The disease was classified in Japan as an intractable disease, with public funding of medical expenses and research [40].

1.2. Moyamoya disease, moyamoya syndrome and moyamoya angiopathy

The Japanese criteria define MMD as a disease of unknown cause and call the identical angiographic picture occurring in a patient with an associated comorbidity "quasi-moyamoya" or moyamoya syndrome [1]. The guideline of the European Stroke Organisation (ESO) uses the term moyamoya angiopathy (MMA) for the whole group and distinguishes idiopathic MMD from MMS associated with hereditary diseases, such as trisomy 21, neurofibromatosis type 1 (NF1) or sickle cell disease, and with acquired conditions, such as head and neck radiotherapy [2]. The list of excluded comorbidities has changed over time: in 2021, atherosclerosis, head trauma and the "others" category were removed, and hyperthyroidism no longer excluded the diagnosis of MMD [1]. The 2023 statement of the American Heart Association/American Stroke Association (AHA/ASA) notes that there is no universal agreement on these exclusions and that the Japanese criteria do not include sickle cell disease, rare in East Asia but recognised in the West as a cause of MMS [18].

The distinction between MMD and MMS is nosological rather than aetiological. The discovery of genes such as RNF213, ACTA2, GUCY1A1, BRCC3, ANO1 and NOS3 showed that part of the so-called idiopathic disease has a monogenic or susceptibility basis, and that some "syndromic" forms (with achalasia, aortopathy or hypogonadism) are Mendelian disorders with the angiopathy as the main manifestation [35, 37, 45]. OMIM organises these forms into a phenotypic series of eight entries (MYMY1 to MYMY8), and Orphanet treats moyamoya angiopathy as a clinical group with monogenic subentities [4, 6].

1.3. Milestones of molecular discovery

Familial clustering was recognised early: about 10% of Japanese cases were familial, 76% of them among siblings [6]. Linkage studies pointed to loci at 3p24-p26 (MYMY1) and 8q23 (MYMY3, 12 Japanese families, LOD 3.6) without identifying the gene [6, 8]. In 2009, Guo and colleagues showed that heterozygous ACTA2 variants, already known as a cause of familial thoracic aortic aneurysm and dissection, also predispose to premature coronary artery disease, early ischaemic stroke and MMD (MYMY5) [46]. In 2010, Milewicz and colleagues described the de novo ACTA2 variant p.Arg179His as the cause of a syndrome of smooth muscle dysfunction throughout the body [47].

In 2011, two Japanese groups identified RNF213 at 17q25.3: Kamada and colleagues, by genome-wide association study (GWAS) with 72 patients and 45 controls, and Liu and colleagues, by linkage in eight three-generation families followed by exome sequencing, both with a founder variant (p.R4810K, historical notation of NM_001256071.3:c.14429G>A p.(Arg4810Lys), designated p.R4859K in an alternative numbering) [48, 49]. In the same year, Miskinyte and colleagues associated Xq28 deletions removing BRCC3 with an X-linked moyamoya syndrome [37]. In 2014, Hervé and colleagues described the autosomal recessive disorder with moyamoya and achalasia caused by loss of function of GUCY1A3 (now GUCY1A1), the α1 subunit of soluble guanylyl cyclase, the main nitric oxide receptor [35]. DIAPH1 (2021), de novo RNF213 variants in children with diffuse occlusive disease (2021), gain-of-function ANO1 (MYMY7, 2023) and homozygous NOS3 (MYMY8, 2023) followed, in addition to de novo variants in CHD4, CNOT3 and SETD5 [13, 20, 38, 45, 50].

1.4. Definition of the entities and scope

Moyamoya angiopathy is treated as a class. MMD is defined as stenosis or occlusion of the arteries centred on the terminal portion of the intracranial ICA, with moyamoya vessels in the vicinity, bilateral or unilateral, after exclusion of the causes of quasi-moyamoya [1]. The monogenic forms covered in depth are: RNF213-associated susceptibility (MYMY2) and the syndrome due to de novo RNF213 variants; ACTA2-related disease (MYMY5) and multisystemic smooth muscle dysfunction syndrome (MSMDS; smooth muscle dysfunction syndrome, SMDS) due to substitutions of arginine residue 179; GUCY1A1-related MYMY6; MYMY4 due to BRCC3 deletion; ANO1-related MYMY7; NOS3-related MYMY8; and the genes with more limited evidence (DIAPH1, IRAG1/MRVI1, CHD4, CNOT3, SETD5). The systemic diseases that cause MMS (NF1, trisomy 21, sickle cell disease, Alagille syndrome, microcephalic osteodysplastic primordial dwarfism type II, SAMHD1-related Aicardi-Goutières syndrome, Grange syndrome and others) are included in the differential diagnosis. Radiation-induced moyamoya syndrome, which is not genetic, is mentioned because NF1 increases its risk [51].

2. Molecular Genetics and Pathophysiology

2.1. Common pathophysiology of the angiopathy

The characteristic lesion is fibrocellular intimal thickening, mainly in the terminal portion of the ICAs, with undulation of the internal elastic lamina, thinning of the media and sometimes lipid deposition; the arteries of the circle of Willis may be narrowed, and numerous small perforating and anastomotic channels form around them [1]. The neointimal cells express smooth muscle cell (SMC) markers without the lipid and inflammatory deposition of atherosclerosis, which separates MMA from atherosclerotic stenoses [45]. Reduction of the outer diameter of the affected arteries (negative remodelling), visible on heavily T2-weighted MRI sequences, is another feature that distinguishes it from atherosclerosis, in which only the lumen narrows [1]. This negative remodelling also explains the failure of angioplasty and stenting [52].

The manifestations result from three mechanisms that balance one another over the course of the disease: haemodynamic ischaemia due to reduced perfusion and cerebrovascular reserve, artery-to-artery embolism (microembolic signals on transcranial Doppler) and haemorrhage from rupture of fragile collaterals, choroidal anastomoses or aneurysms formed on the collaterals [2, 40]. Hyperventilation, dehydration, fever and hypotension precipitate ischaemia because they reduce flow in a collateral-dependent territory [2, 40]. The posterior circulation functions as a collateral pathway; when the posterior cerebral artery (PCA) is involved, the risk of infarction, cognitive impairment and, in adults, posterior haemorrhage from choroidal and thalamoperforating anastomoses increases [2].

The known genes cluster into partially overlapping biological pathways: ubiquitination and cell-autonomous immunity (RNF213, BRCC3), the contractile apparatus and SMC differentiation (ACTA2, DIAPH1), the nitric oxide–soluble guanylyl cyclase–cyclic GMP pathway (NOS3, GUCY1A1, IRAG1), SMC ion channels (ANO1) and chromatin remodelling (CHD4, CNOT3, SETD5, as well as YY1AP1 and SMARCAL1 in syndromic forms) [13, 18, 35, 38, 45, 50].

2.2. Table of genes and types

Sources for the table: [2, 6–13, 20, 21, 35, 37, 38, 45, 46, 50, 53, 54].

2.3. RNF213 (MYMY2)

RNF213 (mysterin) encodes a cytosolic protein of about 591 kDa with AAA+ ATPase domains (Walker motif) and a RING domain with E3 ubiquitin ligase activity [49, 55]. The protein ubiquitinates proteins and lipids, including bacterial lipopolysaccharide, and acts as a sensor of cell-autonomous immunity; it is also involved in lipid storage and in angiogenesis through non-canonical Wnt signalling [18, 55]. The 2023 AHA/ASA statement considers that p.R4810K may be hypomorphic and fail to degrade two substrates, NFAT1 and filamin A, and notes that most disease-associated variants are missense and located in the C-terminal region, suggesting a dominant negative or gain-of-function effect rather than haploinsufficiency [18]. In the Liu study, the mutant proteins p.(Arg4810Lys) and p.(Asp4013Asn) (p.R4810K and p.D4013N in the article's one-letter notation) had normal expression, localisation and ubiquitination activity, but knockdown of rnf213 in zebrafish produced irregular walls in the trunk arteries and abnormal vascular sprouting [7, 49]. The hypothesis of an immunological or infectious "second hit", supported by the antimicrobial function of the protein and by cases of MMA after infection, remains under study [18].

The East Asian founder variant is designated, on the MANE Select transcript, NM_001256071.3:c.14429G>A p.(Arg4810Lys), corresponds to NC_000017.11:g.80385145G>A (GRCh38) and to dbSNP rs112735431, and is recorded in OMIM as allele 613768.0002 [7, 56, 57]. The early Japanese literature used another numbering, c.14576G>A p.(Arg4859Lys), referring to a longer transcript (OMIM allele 613768.0001); the two designations describe the same substitution, and VariantValidator rejects coordinate c.14576 on the current version NM_020954.4, so the conversion should not be done by hand [19, 48, 56]. In ClinVar, the variant (VCV000039700) has conflicting classifications, last evaluated on 15 June 2026, consistent with its status as a susceptibility allele rather than a Mendelian pathogenic variant [57].

The association studies gave the following values. Kamada and colleagues found the variant in 95% of families, 73% of non-familial cases and 1.4% of Japanese controls (OR 190.8; 95% CI 71.7–507.9); in the families, it was present in 39 of 41 affected individuals (92.9%), but also in 13 of 28 unaffected relatives (46.4%), and none of the 15 relatives without the variant had the disease [7, 48]. Liu and colleagues, in 251 cases and 707 East Asian controls, obtained an OR of 111.8 and a founder haplotype shared by 42 families [49]. Miyatake and colleagues found it in 95.1% of familial cases, 79.2% of sporadic cases and 1.8% of controls (OR 259) [19]. The carrier frequency in the general Japanese, Korean and Chinese populations lies between 0.5% and 2.5%, and the variant is absent in the European patients and controls studied [7, 20, 55, 58].

In non-Asian patients, the role of the gene is expressed through different rare variants. In a genetic burden study with 68 European probands and 573 French controls, rare RNF213 missense variants were associated with MMA (OR 2.24; 95% CI 1.19–4.11), with a greater effect in childhood-onset or familial cases (OR 4.54; 95% CI 1.80–11.34); these variants had higher pathogenicity scores and clustered in a C-terminal hotspot that includes the RING domain (amino acids around 3997–4035), unlike the variants in controls, which were distributed along the gene [20, 58]. The p.(Asp4013Asn) variant segregated with MMD in two large European families [58]. Liu and colleagues had already described p.(Asn3962Asp), p.(Asp4013Asn), p.(Arg4062Gln) and p.(Pro4608Ser) in Caucasian cases, and p.(Asp4863Asn), p.(Glu4950Asp), p.(Ala5021Val), p.(Asp5160Glu) and p.(Glu5176Gly) in Asians (in the one-letter notation of the original publications: N3962D, D4013N, R4062Q, P4608S, D4863N, E4950D, A5021V, D5160E and E5176G) [49]. Revalidation of these designations in VariantValidator and ClinVar on 1 November 2026 confirmed the numbering of all of them, but showed that p.(Ala5021Val) is now classified as benign or likely benign in ClinVar, so the list corresponds to the variants described in those cohorts and not to a set of confirmed causal variants [56]. These protein positions, like those of the de novo variants described below, follow the numbering of transcript NM_001256071 (version .2 in the European study), the same in which the founder variant occupies residue 4810 [56, 58].

A third mechanism is the de novo variant. Pinard and colleagues, in 28 trios of children with MMD and unaffected parents, identified three de novo RNF213 variants (one in the RING domain and two in a conserved distal region, amino acids 4114–4120) and gathered two further published cases in the same region: all had aggressive MMD before 3 years of age and occlusions in other arteries (renal, abdominal aorta, iliac, femoral), which defines a distinct syndrome [20]. The de novo variants described include p.(His4014Asn) and p.(Cys4017Ser), which affect cysteine or histidine residues that coordinate zinc in the RING domain, and p.(Lys4115del), p.(Ser4118Phe) and p.(Val4146Ala) [58]. The ESO guideline adds that these de novo variants in the E3 domain are associated, in infants, with hepatic, renal and cutaneous manifestations that delay diagnosis [2].

2.4. Penetrance and modifiers of RNF213

The penetrance of p.Arg4810Lys in the heterozygous state is low at the population level, estimated at about 1 in 150 Japanese carriers (0.67%), with ranges cited between 1/150 and 1/300; in the homozygous state it is above 78% [18–20]. In a familial context penetrance is much higher: in European families with rare variants, about 25% of carriers were affected, which led to the oligogenic hypothesis (additional factors shared within the family) [58]. The AHA/ASA paediatric statement cites a familial penetrance of about 23% [36]. In highly clustered Japanese families, transmission is predominantly maternal and affected mothers have more affected female offspring, suggesting imprinting; in a Korean population-based study, the risk in first-degree relatives was highest with an affected twin, followed by a sibling, the mother and, lastly, the father [18].

2.5. ACTA2 (MYMY5 and MSMDS)

ACTA2 encodes the vascular SMC-specific isoform of α-actin, a central component of the contractile apparatus [46]. Causal variants are almost all missense with a dominant negative effect on filament polymerisation and force transmission [59]. In the series of Guo and colleagues (20 families), only half of carriers had aortic disease, and carriers also had premature coronary artery disease and early ischaemic strokes, including MMD; the diseased arteries and explanted SMCs showed increased proliferation, which would contribute to the occlusive lesions [46]. In three families with thoracic aortic aneurysm and MMD, stroke occurred between 5 and 46 years of age, and one member of each family had isolated moyamoya; the variants included p.(Arg258His) and p.(Arg258Cys), in the protein numbering of transcript NM_001613.4 [10, 60]. Roder and colleagues found ACTA2 p.(Arg179His) in 1 of 39 Europeans with sporadic MMD (stroke at 3 years) [10].

Substitutions of arginine residue 179, mainly NM_001613.4:c.536G>A p.(Arg179His) (NC_000010.11:g.88941309C>T), broadly abolish SMC differentiation and cause MSMDS; with a single exception, they arose de novo [21, 59, 60]. The residue lies near a protein-protein interaction site on the actin surface; sequence analysis found no mutable motif to explain the recurrence, which may reflect preferential ascertainment of a highly recognisable phenotype [39]. In a mouse model with inducible insertion of Arg179Cys into one Acta2 allele in SMCs, resulting in 67% mutant SMCs in the aorta, Acta2 R179C/+ SMCs explanted from the ascending aorta have reduced levels of smooth muscle myosin heavy chain, α-actin, transgelin and calponin-1, less α-actin filament formation, increased proliferation and migration, and retain the metabolic profile of stem cells: a higher extracellular acidification rate, indicative of glycolysis, and lower basal, ATP-linked and maximal oxygen consumption [61]. The oxidative defect corresponds to reduced expression of 8 of the 13 electron transport chain subunits encoded in mitochondrial DNA, lower levels of PGC-1α and TFAM and reduced quantity and activity of complexes I and IV, with no differences in mitochondrial number, area or cristae on electron microscopy, indicating mitochondrial dysfunction rather than reduced mitochondrial mass [61]. Knockdown of Cox4i1 by RNA interference in normal SMCs lowered the expression of Acta2, Cnn1, Tagln and Myh11, linking respiratory chain function to SMC differentiation [61]. After left carotid ligation at 8 weeks of age, 22% of mutant mice died by the 5th postoperative day (11 of 52, versus 0 of 40 controls; p<0.01), with higher mortality in females (8 of 28; p=0.012) than in males (3 of 24; p=0.09), a gradient that parallels the female predominance of cerebrovascular disease in MSMDS; on day 21, survivors had carotid lumens filled with smooth muscle α-actin-positive cells, loss of pyramidal neurons in the CA2-CA3 regions of the hippocampus and in the striatum, a narrowed circle of Willis and more dilated and tortuous ipsilateral leptomeningeal collaterals, whereas controls kept patent lumens and had no neuronal loss [61].

In aortic disease, variants affecting p.Arg179 and p.Arg258 confer a significantly higher risk of an aortic event, and p.Arg185Gln and p.Arg118Gln a lower risk; in forms that predispose to occlusive disease, screening for coronary and cerebrovascular disease is recommended [39, 59]. ACTA2-related aortopathy belongs to the group of heritable aortopathies.

2.6. The nitric oxide pathway: GUCY1A1, NOS3 and IRAG1

GUCY1A1 (formerly GUCY1A3) encodes the α1 subunit of soluble guanylyl cyclase (sGC), an α1β1 heterodimer that is the receptor for nitric oxide (NO) and produces cyclic GMP (cGMP), a second messenger that relaxes vascular and oesophageal SMCs and regulates vascular tone and remodelling [35, 62]. In the three initial consanguineous families (France and Algeria), the homozygous variants were, on transcript NM_000856.4 and with the same numbering on the MANE Select transcript NM_001130682.3, c.1045C>T p.(Arg349Ter), c.1170del p.(Glu391LysfsTer19) and c.1086+1G>A (effect on the protein not determined), all with predicted premature termination [35, 63]. The patients' platelets lacked the α1 and β1 subunits, aggregated and adhered less to collagen and von Willebrand factor and did not phosphorylate VASP in response to NO donors, which revealed a stimulatory role of sGC in platelet activation [35]. The authors proposed that loss of the NO–sGC–cGMP pathway alters vascular remodelling at sites of disturbed laminar flow, such as the ICA bifurcation, and that achalasia results from lack of NO-mediated relaxation of the lower oesophageal sphincter [11, 35].

The series was extended by Wallace and colleagues, who found compound heterozygous loss-of-function variants in 2 of 96 European probands with MMD, one of them without achalasia at 3 years of age, and estimated that GUCY1A1 may account for up to 2% of MMD in patients of European origin; Guey and colleagues described a Turkish woman homozygous for p.(Arg593His), with symptoms at 43 years, stenosis of the superior mesenteric artery ostium and no achalasia [11]. The heterozygous loss-of-function variant in the same gene, in a family with myocardial infarction, and the common variants associated with coronary artery disease (OR 1.08) show that GUCY1A1 has dose-dependent and tissue-dependent effects [62].

NOS3 encodes endothelial NO synthase. In two consanguineous families (Morocco and Tunisia), homozygous variants absent from population databases caused severe infantile moyamoya with steno-occlusive PCA lesions (MYMY8) [13]. IRAG1 (MRVI1) encodes the inositol trisphosphate receptor-associated cGMP-dependent kinase substrate, which regulates intracellular calcium release; the p.(Pro186Ser) polymorphism (rs35857561) segregated with moyamoya in two European families with NF1, suggesting a modifier of MMS in NF1 [54, 64].

2.7. BRCC3 (MYMY4)

BRCC3 encodes a deubiquitinase of lysine 63-linked ubiquitin chains, a subunit of the BRCA1-A and BRISC complexes [37, 65]. In three families with X-linked moyamoya syndrome, the Xq28 deletions overlapped in a 3.4 kb critical region that includes exon 1 of MTCP1/MTCP1NB and exons 1–3 of BRCC3, with loss of expression of both; zebrafish with brcc3 knockdown had angiogenesis defects corrected by endothelial expression of the gene [9, 37]. The penetrance of the angiopathy in males with this deletion is above 75%, the highest among Mendelian disorders with moyamoya [58].

2.8. ANO1 (MYMY7)

ANO1 encodes anoctamin-1, a calcium-activated chloride channel expressed in SMCs. Pinard and colleagues identified the p.(Met658Val) variant (NM_018043.6) in two related families (LOD 3.3) and six other rare variants; in patch-clamp assays, p.(Met658Val), p.(Thr740Ile) and p.(Glu459Lys) increased sensitivity to intracellular calcium (gain of function), p.(Arg890Gln) reduced function and p.(Arg77Gln) and p.(Ser196Thr) did not differ from wild type [12, 38]. Transmission was dominant with incomplete penetrance in six families and recessive in one consanguineous family homozygous for p.(Glu170Lys); in one family, carriers who inherited the variant from their mother became ill earlier, suggesting paternal imprinting [12].

2.9. DIAPH1, CHD4, CNOT3 and SETD5

DIAPH1 encodes the formin diaphanous-1, a Rho effector that forms unbranched actin filaments in vascular cells and platelets. Of 24 probands (75% European) studied as trios, three had DIAPH1 variants (two de novo and one damaging inherited variant); in a validation cohort of 84 probands, three more patients had ultra-rare variants; all six were non-Asian, had severe bilateral MMD and transfusion-dependent thrombocytopenia, as did one patient with compound heterozygous variants in the interactor EVL [50]. The same gene causes autosomal dominant low-frequency deafness (DFNA1), macrothrombocytopenia with deafness and a recessive syndrome of seizures, cortical blindness and microcephaly [66]. Analysis of 39 trios showed de novo variants in CHD4, CNOT3 and SETD5, neurodevelopmental disorder genes, and additional rare variants in 158 probands supported CHD4 and CNOT3; for example, de novo CNOT3 p.(Gln215Ter) in a woman with intellectual disability and bilateral MMA diagnosed at 20 years [45]. In the GeneReviews chapter on CHD4-related neurodevelopmental disorder, two of three children with postnatal stroke had MMD [67].

2.10. Inheritance, penetrance and expressivity

The mode of transmission of idiopathic MMD is unknown; the data favour autosomal dominant inheritance with incomplete penetrance and the influence of other genetic and environmental factors [18]. About 10–12% of Japanese patients have a family history [6, 30], 7–12% of probands in North American and European cohorts have affected relatives [38], and familial clustering is rare in white patients (3.4% in a large North American series) [18, 36]. Orphanet lists autosomal dominant, autosomal recessive, multigenic/multifactorial and X-linked recessive inheritance for MMD, reflecting the heterogeneity; the GeneReviews chapter on generalised arterial calcification of infancy attributes both patterns, dominant and recessive, to RNF213 [3, 68]. Expressivity varies within the same family: with GUCY1A1, achalasia was constant and the angiopathy appeared in about half of carriers; with ANO1, a carrier mother had only multiple intracranial aneurysms; with ACTA2, some relatives had only aortic aneurysm and others only moyamoya [10, 12, 35].

3. Clinical Spectrum and Manifestations by System

3.1. Data sources and what they measure

The clinical data come from Japanese national surveys, from registries and administrative databases (Korea, USA), from hospital cohorts of referral centres (Boston, Zurich, Paris, Houston) and from small genetic series. Hospital series over-represent operated and symptomatic patients; administrative databases rely on diagnostic codes not validated to distinguish MMD from MMS [18]; and monogenic series often describe fewer than ten families, so the frequencies by gene should be read as descriptions and not as risk estimates. The natural history without surgery is known only from retrospective studies and conservative arms, because revascularisation is standard practice in symptomatic disease [52].

Age at onset has a bimodal distribution, with one peak in childhood, at around 10 years, and another at around 40 years [18]. In the 2003 Japanese survey, the peaks were at 10–14 years in males and 20–24 years in females [30]; OMIM records a smaller peak in the fourth decade [6]; and several Chinese epidemiological studies place the paediatric peak between 5 and 9 years, with variable adult peaks (40–44 years in Taiwan; 35–39 years in Nanjing and the central plain) [34].

3.2. Presentation in children

In children, ischaemia predominates. In a single-centre cohort of 54 children with moyamoya arteriopathy (median 7.5 years at diagnosis), 70% had TIAs and 48% acute ischaemic stroke; the most frequent deficits were hemiparesis or hemihypoaesthesia (72%), with chronic headache in 52% and seizures in less than 10%, and 52% already had silent infarcts on the first imaging study [36]. In another cohort of 60 children presenting with TIA, 92% had recurrent TIAs and 23% progressed to infarction; the severity of the arteriopathy at diagnosis corresponded to the risk of subsequent stroke [36]. Moyamoya is associated with 6–10% of strokes and TIAs in childhood, and progressive arteriopathies have the highest risk of recurrence (1 in 3 children with a new stroke in the first year) [36]. Haemorrhage is rare in children (about 3% versus 25–60% in adults) [2]. In all regions studied, more than 75% of patients with childhood onset present with ischaemia [69].

Episodes are classically triggered by hyperventilation, as in crying, blowing (wind instruments, balloons) or eating hot soups and noodles, and by dehydration during vomiting and diarrhoea [40]. Cognitive impairment, involuntary movement disorders, migraine-like headaches and seizures are part of the spectrum and may be the only manifestation [2, 34]. PCA involvement, observed in 35% of a mainly European paediatric cohort, and onset before 2 years of age are associated with a greater stroke burden, worse non-verbal IQ, lower processing speed and worse social outcomes 10 years after surgery [2].

3.3. Presentation in adults

In adults, ischaemia and haemorrhage coexist. The proportion of haemorrhage as the first manifestation varies between regions: 56% in China, 52% in Taiwan, 29% in Hawaii, 21% in Japan and 10% in Iowa [69]. In a cohort of 200 European Caucasian patients, 9.5% had a haemorrhagic presentation [2]. Haemorrhage results from rupture of the dilated collaterals or of peripheral aneurysms formed on them; posterior haemorrhages, attributed to the choroidal and thalamoperforating arteries, recur more often [18, 40]. In the Japan Adult Moyamoya (JAM) trial, the non-surgical group had rebleeding in 17.1% per year when the initial haemorrhage was posterior, and the presence of choroidal anastomosis raised rebleeding to 13.1% per year, versus 0.3% per year without it [2, 40]. Vascular cognitive dysfunction, headaches and psychiatric disorders often accompany adult disease, and have been associated with regional hypoperfusion measured by SPECT and PET [2].

3.4. Natural history and progression

The angiopathy is progressive in both age groups, faster in children. Without surgery, the risk of stroke in a symptomatic child has been estimated at 32% at 1 year and 66–90% at 5 years, falling to less than 5% at 1 and 5 years after revascularisation in most series [36]. In adults, retrospective series from the USA and Germany recorded 80–82% cerebrovascular recurrence at 5 years after a first stroke, whereas a prospective study of 49 adults recorded 10.2% over 3.7 years [2]. In a prospective study of 90 multi-ethnic patients under conservative management (42.8 months), impaired cerebrovascular reserve, East Asian origin and previous TIA were independent predictors, and the annual risk of stroke or silent lesion ranged from 0.5% to 20% depending on these factors [2].

Initially unilateral disease progresses to bilateral in 59% of Asian children, in about 30% of Caucasian children over 2.2 years and in 38.9% of Caucasian adults over 12.7 months; in operated and non-operated Asian adults, progression was 20% over 73.6 months, and more than half of progressions coincided with a stroke [2]. Progression to the PCA may appear 0.5 to 15 years after surgery (mean 5.4 ± 4.4 years) [2]. In asymptomatic patients, a Japanese survey (13–59 years, mean follow-up 44 months) recorded cerebrovascular events, including TIA, in 21% and an annual risk of ischaemic or haemorrhagic stroke of 3.2%; the 5-year interim analysis of the Japanese AMORE registry recorded stroke in 7 of 103 asymptomatic patients (6 haemorrhagic), an annual risk of 1.4% per person and 1.0% per hemisphere, with microbleeds and developed choroidal channels as predictors of haemorrhage [2, 52]. In Korea, survival at 1 and 5 years was 96.9% and 92.9% in adults and 99.6% and 99.3% in children [31].

3.5. RNF213-associated disease

Carriers of p.Arg4810Lys show more familial clustering, earlier onset, ischaemic presentation and PCA involvement [55]. Homozygotes have a median onset at 3 years (versus 7 in heterozygotes and 8 in non-carriers); 60% are diagnosed before 4 years of age and all present with infarction [19]. In another study, homozygotes more often had a family history, onset before 5 years, infarction at diagnosis and cognitive impairment at 1 year [2]. In the heterozygous state, the variant was not associated with worse outcome in three cohorts of operated patients (129 Japanese children, 94 Japanese patients and 2545 Chinese patients) [2].

RNF213-related vasculopathy extends beyond the cerebral arteries. p.Arg4810Lys is present in about 20–22% of Japanese patients with stenosis or occlusion of large intracranial arteries without moyamoya and is associated with intracranial atherosclerosis (OR 5.19) [2, 20, 55]. Homozygotes may have multiple stenoses of extracranial arteries and peripheral pulmonary artery stenosis, even without MMD [68]. De novo variants at amino acids 4114–4120 and in the RING domain cause MMD before 3 years of age with stenosis of the renal arteries, abdominal aorta and iliac and femoral arteries [20], and some present with hepatic, renal and cutaneous involvement [2]. The GeneReviews chapter on generalised arterial calcification of infancy also mentions a Grange syndrome-like phenotype (renal stenosis, brachydactyly, syndactyly of the toes and intellectual disability) with a pathogenic RNF213 variant [68].

3.6. ACTA2-associated disease and MSMDS

In carriers of ACTA2 variants in general, aortic disease is the best-known manifestation (aortic events in 48%, 88% of them dissections, with 25% mortality; cumulative risk of an aortic event by 85 years of 76%), but there is also premature coronary artery disease, early ischaemic stroke and moyamoya, livedo reticularis and iris flocculi [39, 59]. Arg179-related MSMDS is a congenital multisystem disorder. In the series of 33 patients of Regalado and colleagues (median 12 years), all had congenital mydriasis or related pupillary anomalies (fixed pupils, iris hypoplasia, partial aniridia) and patent ductus arteriosus or aortopulmonary window with neonatal intervention; 36% underwent surgery for aortic aneurysm or dissection (median 14 years), aortic disease was fully penetrant by 25 years, 9% had axillary aneurysms with thromboembolism and 9 patients (27%) died between 6 months and 32 years of age, from aortic dissection, postoperative complications, stroke or pulmonary causes [21, 39].

The cerebrovascular disease of MSMDS differs from idiopathic MMD: the proximal ICAs are dilated, the intracranial arteries have an abnormally straight course, the stenoses affect the distal ICA, the cerebral arteries and the posterior circulation, and the basal collateral network is absent [21, 39]. The chapter on heritable aortopathies in Clinical Cardiogenetics, which counted about 40 described patients with MSMDS, describes this involvement as resembling moyamoya disease without being entirely similar to it and illustrates it with the MR angiogram of a 17-year-old girl (figure 19.13 of the book): left ICA dilated up to its terminal portion, occlusive disease of the distal intracranial circulation, abnormally straight course of the intracranial arteries and absence of moyamoya collaterals [39]. Intracranial stenosis or occlusion was present in 77% of patients (median 8 years, as early as 5 days of life), affected more than one artery in 53% (distal ICA 53%, cerebral arteries 60%, vertebral 20%, basilar 17%), and 95% had periventricular white matter hyperintensities; ischaemic stroke occurred in 27% and seizures in 18% [21]. In a study of 113 MRIs from 27 patients, infants had acute internal border-zone white matter infarcts, the terminal ICA narrowed progressively in 80% and this narrowing correlated with critical stenoses and infarcts, whereas the white matter hyperintensities remained stable and the cystic lesions increased over time [53]. Other manifestations include pulmonary hypertension (with bilateral lung transplantation at 18 months in one case), asthma and cystic lung disease, hypoperistalsis and intestinal malrotation, hypotonic bladder with megacystis, cryptorchidism, atonic uterus in pregnancy, tortuosity of the retinal arteries and dysautonomia with difficulty regulating blood pressure [21, 39, 53, 70]. The Oxford Desk Reference includes MSMDS in the spectrum of visceral myopathies, alongside megacystis-microcolon-intestinal hypoperistalsis syndrome (heterozygous ACTG2 or biallelic MYH11) and chronic intestinal pseudo-obstruction, which may manifest in pregnancy as fetal megacystis [71].

3.7. MYMY6 (GUCY1A1)

Achalasia with neonatal onset or in the first months of life is the constant feature: in the three initial families it was present in 9 of 9 patients, with Heller myotomy between 2 months and 3 years of age, and severe intracranial angiopathy appeared in 4 of 9 (bilateral moyamoya in 3) [35]. The most severely affected had infarcts in early childhood with hemiparesis and disability, with involvement of the anterior and posterior circulations; the less severely affected had normal brain imaging but other vascular abnormalities: hypertension in 4 patients (one with malignant hypertension, retinopathy, ischaemic optic neuropathy and renal failure at 25 years), Raynaud phenomenon, livedo reticularis and erectile dysfunction [11, 35]. There were no clinical haemorrhages despite the platelet dysfunction [35]. The European patients of Wallace and colleagues showed that achalasia may be absent: one had achalasia at 4 years and stroke at 18, another had a severe stroke at 20 months, hypertension and mild dilatation of the ascending aorta without achalasia at 3 years [11].

3.8. MYMY4 (BRCC3), MYMY7 (ANO1) and MYMY8 (NOS3)

In the Xq28 deletion syndrome, the nine males described in three families had acute neurological events between 4 and 36 years of age, ischaemic and haemorrhagic; all had short stature and facial dysmorphism (broad nose, deep-set eyes, low-set ears, hypertelorism, ptosis, long philtrum, anteverted nares), and other features were hypergonadotropic hypogonadism with azoospermia (7/9), premature greying (6/9), partial growth hormone deficiency (4/9), hypertension (3/9), dilated cardiomyopathy (3/9), premature coronary artery disease (1/9), early cataracts (4 patients from one family) and developmental delay (2) [9, 37]. The Algerian proband died at 34 years from heart failure [9].

In MYMY7, the 11 patients from 7 families had onset between 9 months and 50 years of age, with hemiparesis, poor school performance, choreiform movements, migraine and TIAs; the disease included intracranial aneurysms (including of the basilar tip), progressive basilar occlusion and posterior collaterals, and one carrier had only multiple aneurysms [12, 38]. In MYMY8, a Moroccan woman had transient hemipareses from 9 months to 9 years of age with infarcts in the PCA and MCA territories, and a Tunisian boy with intrauterine growth restriction, short stature and dysmorphic features had strokes at 6 and 12 months (the second during an orchidopexy) and new strokes during respiratory infections despite two revascularisations [13].

3.9. Moyamoya syndromes in systemic genetic diseases

The probability of MMA varies widely between Mendelian disorders, from less than 2% in the RASopathies to more than 75% in BRCC3 deletion [58]. In trisomy 21, the probability of moyamoya is 26 times that of the general population: in a US national inpatient sample (2002–2009), 3.8% of patients with MMD had trisomy 21 (9.5% of those under 15 years), with more ischaemia (15.3%) than haemorrhage (2.7%) [36, 72]. In NF1, the prevalence of moyamoya is about 2–5%, the vasculopathy (intimal and medial hyperplasia) may also affect the renal arteries, the mid-aorta and the coronary arteries, and moyamoya occurs about three times more often than expected after cranial irradiation [25, 36, 73]. In sickle cell disease, up to 50% of children with HbSS or HbSβ0 have cerebrovascular disease by 14 years of age, and among children with stroke on chronic transfusion, moyamoya collaterals were present in 43% and were associated with more recurrent events (58% versus 28%) [27, 74].

In microcephalic osteodysplastic primordial dwarfism type II (MOPD II, PCNT), 22 of 47 patients had moyamoya vasculopathy and 25 intracranial aneurysms, about half had a stroke, and the risk of moyamoya was concentrated before 5 years of age, from intrauterine life onwards [23, 75, 76]. In Alagille syndrome, vascular abnormalities occur in 15–30%, neurovascular events affect up to 15% and accounted for 34% of mortality in a large series, and moyamoya may be present in up to one-third of patients [26, 77]. In Portugal, a group from Centro Hospitalar de São João (Porto) described a girl with posterior embryotoxon, peripheral pulmonary stenosis and typical facies, without cholestasis or vertebral anomalies, with a moyamoya pattern on MR angiography and the heterozygous JAG1 duplication NM_000214.3:c.715dup, predicted as p.(Ser239LysfsTer3) [78, 79]. In PHACE syndrome (posterior fossa malformations, haemangioma, arterial lesions, cardiac anomalies or coarctation of the aorta and eye anomalies), a definitive diagnosis requires a segmental facial infantile haemangioma (usually larger than 5 cm) and two major criteria; the cervical and intracranial arterial lesions include moyamoya-like forms, and the AHA/ASA paediatric statement calls for brain MRI, MR angiography of the head and neck, echocardiography and ophthalmological examination, with vascular surveillance for intermediate- or high-risk lesions [36]. In Aicardi-Goutières syndrome, large intracranial vessel disease (stenoses, moyamoya pattern, aneurysms) is almost exclusive to biallelic SAMHD1 [24, 80]. Cases have been described in Schimke immuno-osseous dysplasia (SMARCAL1), Fanconi anaemia, ESCO2 spectrum disorder (moyamoya in 2%), CHOPS syndrome (AFF4), Smith-Magenis syndrome with a large deletion, Legius syndrome, osteopathia striata with cranial sclerosis and Grange syndrome (YY1AP1) [28, 29, 68, 81–87].

Among acquired causes, cranial irradiation for childhood brain tumour caused moyamoya in 3.5% of 345 children, sooner with NF1 (median 38 versus 55 months) and with doses above 5000 cGy; each 100 cGy increased the risk by 7% [51]. The AHA/ASA cites rates of 2–4.3% after irradiation and up to 60% in NF1 irradiated for optic pathway glioma [18]. In the USA, white patients with moyamoya have more autoimmune diseases, such as type 1 diabetes (8.5% versus 0.4% in the population) and thyroid disease (17% versus 8%) [36], and hyperthyroidism was found in 10.5% of 114 Japanese children with MMD versus 0.9% of controls [1].

3.10. Multisystemic smooth muscle dysfunction and the related ACTA2 spectrum

Designation and genotypes. MSMDS (also termed SMDS, smooth muscle dysfunction syndrome) was described in 2010 and results from heterozygous ACTA2 missense variants that replace arginine 179 with one of five amino acids: histidine, NM_001613.4:c.536G>A p.(Arg179His); cysteine, c.535C>T p.(Arg179Cys); leucine, c.536G>T p.(Arg179Leu); serine, c.535C>A p.(Arg179Ser); and glycine, c.535C>G p.(Arg179Gly) [43, 88]. Substitution by proline is possible under the genetic code but has not been observed [43]. The five substitutions are associated with similar age at onset and manifestations [43]. In a compilation of 90 patients published up to 2024, the genotype was described in 85, and p.(Arg179His) accounted for 63.5% of cases [42]. The same review reports 120 ACTA2 variants recorded in HGMD in April 2024 (108 missense or nonsense, 5 splicing and 7 microdeletions) [42].

Pooled data from published reports. Of the 90 patients, 80.0% came from reports from Europe and the USA, 7 from China and 1 from Japan, and 66.7% were female. Onset ranged from fetal life to adulthood. Congenital mydriasis (92.6%) and patent ductus arteriosus (91.6%) were the most frequent manifestations. Interstitial lung disease occurred in 12 patients (13.3%), all with onset in the first year of life, 11 with p.(Arg179His) and 1 with p.(Arg179Cys) [42]. Among the patients with p.(Arg179His), 10 (18.5%) had died at the time of publication, between 0.5 and 32.7 years of age. Only 6 (11.1%) had follow-up described, and 38 (70.4%) had no prognostic information [42]. These are case report data, subject to publication and ascertainment bias, and do not estimate prevalence or lifetime risk. The series of Regalado and colleagues (33 patients, section 3.6) remains the source of organ-specific frequencies with a defined denominator [21]. In a study of 24 patients focused on the digestive tract, 75% needed medication for chronic constipation [89]. In a newborn from Rome, the diagnosis of MSMDS with p.(Arg179His) was made in the neonatal period. In the review accompanying the case, the authors include premature coronary artery disease, chronic interstitial lung disease and gallbladder hydrops in the spectrum [90]. A 3-year-old girl with MSMDS had communicating hydrocephalus with intracranial hypertension, the first description of this association. The authors propose looking for ventriculomegaly and signs of raised intracranial pressure in these patients [91].

Cerebral arteriopathy. A review of 15 articles (58 patients with cerebral arteriopathy and a confirmed ACTA2 variant) characterised the internal carotid circulation by dilatation of the proximal segments, distal occlusive disease and dolichoectasia, with ischaemic stroke in childhood. The same review proposes assessment by a geneticist and ACTA2 genotyping in patients diagnosed with MMD and smooth muscle organ dysfunction [92]. The classic description without a basal collateral network was revisited in a longitudinal digital subtraction angiography study of 10 children from Zurich with a confirmed ACTA2 variant. The calibre of the petrous ICA was 8.4 ± 2.8 mm (normative value 3.8 ± 0.4) and that of the supraclinoid ICA 2.6 ± 0.7 mm (normative 3.3 ± 0.3). Classic moyamoya-type collaterals were present in 2 patients at baseline and in 5 at follow-up [93]. In a German family with ACTA2 NM_001613.4:c.592C>T p.(Arg198Cys), the four carriers had straight-coursed intracranial arteries, and none had dilatation of the proximal ICA. The proband also had a basal collateral network, less developed than in idiopathic MMA [88, 94]. The straight course of the intracranial arteries is, according to these authors, the most constant finding of ACTA2-related arteriopathy [94].

Related spectrum outside residue 179. Kaw and colleagues described five patients of European ancestry (Montalcino consortium registry, Houston) with novel variants [43, 88]:

The authors point to other partially overlapping phenotypes at residues Arg258 (ductus arteriosus of variable penetrance, later-onset aortopathy, distal ICA stenosis, white matter hyperintensities and intracranial straightening, without bladder or intestinal complications), Asn117 (moyamoya-like cerebrovascular disease, ductus arteriosus, poorly reactive pupils and retinal artery tortuosity, without aortopathy in one family), Asp181 (arteriopathy, periventricular hyperintensities, hypoplasia of the corpus callosum and ductus arteriosus) and Lys328 (cerebral aneurysms, aortopathy before 50 years, distal ICA stenosis and congenital mydriasis) [43]. They propose the designation "ACTA2-related SMDS" for patients with the complete picture or with variants already associated with the syndrome, and "ACTA2-related SMDS-like" for those with only some of the complications [43]. In infancy, congenital mydriasis, ductus arteriosus and intestinal and bladder complications are present in all patients with a residue 179 variant and are the clinical predictors of the diagnosis [43]. One patient with somatic mosaicism for a variant at residue Asn117 was diagnosed at 62 years and died at 66 [42].

4. Clinical Recognition and Warning Signs

4.1. When to suspect

In children, moyamoya should be considered in the presence of recurrent TIAs with alternating or fixed hemiparesis, particularly when triggered by crying, exertion, fever, heat, hot meals or wind instruments; ischaemic stroke without a cardiac cause; focal seizures, involuntary movements, persistent headaches or unexplained school decline; and silent infarcts found on imaging performed for another reason [23, 36, 40]. Moyamoya is part of the differential diagnosis of all stroke-like episodes in patients with a suspected inherited metabolic disorder, alongside MELAS, the homocystinurias, congenital disorders of glycosylation, Fabry disease and Menkes disease [95–97]. In adults, it should be sought in deep intraventricular or intraparenchymal haemorrhages without hypertension or vascular malformation, and in ischaemic stroke or cognitive impairment in a young adult without vascular risk factors [2, 18].

Targeted investigation for moyamoya is justified in at-risk populations, even without symptoms, because the disease is often recognised before fixed deficits in children with NF1 investigated for other manifestations [98]. The AHA/ASA stresses the risk of attributing a stroke in a child with trisomy 21 and heart disease to another cause when the cause is moyamoya [36].

4.2. Clues to a monogenic cause

4.3. Warning and emergency signs

Emergencies include acute or fluctuating focal neurological deficit, sudden headache with altered consciousness (haemorrhage), seizures and neurological worsening during fever, dehydration, hypotension or anaesthesia [2, 21]. In MSMDS, any sudden change in neurological function should be assessed immediately, and febrile infections with hypoxaemia increase cerebral metabolic demand in a poorly perfused territory [21]. In a child with known moyamoya, hypotension during acute stroke should be avoided and actively treated [36].

4.4. Newborn screening and screening of at-risk populations

There is no newborn screening. Congenital fixed mydriasis in a newborn with patent ductus arteriosus should prompt genetic assessment and ACTA2 testing [21]. In MOPD II, MRI with MR angiography is recommended at diagnosis, including in the neonatal period, because stroke occurs in utero and moyamoya can be present neonatally [23]. In sickle cell disease, annual transcranial Doppler in all children with HbSS and HbSβ0 screens for stroke risk, and the presence of MMS on MR angiography changes the primary prevention strategy [27, 36].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. The 2021 Japanese diagnostic criteria

Source: [1].

Reduction of the outer diameter becomes apparent only from Suzuki stage 3; in stages 1 and 2, MRI may not fulfil all three criteria and angiography is needed [1]. The ESO, for study inclusion, used the 1997 angiographic criteria, with bilateral disease only, and considered unilateral disease only in syndromic cases; the AHA/ASA accepts unilateral disease and retains angiography as the reference standard [52]. The Suzuki angiographic stages (I, narrowing of the carotid bifurcation; II, initiation of the moyamoya network with dilatation of the cerebral arteries; III, intensification of the network with disappearance of the MCA and ACA; IV to VI, minimisation and disappearance of the network with perfusion from the external carotid) describe the course but have limited application to a single angiogram [18]. The Chinese consensus adds the Berlin grading system (haemorrhage, infarction, atrophy and cerebrovascular reserve) and a classification of collaterals that includes the posterior circulation [34].

5.2. Imaging and haemodynamic assessment

MRI with MR angiography is the initial investigation, as it avoids radiation and contrast; FLAIR shows chronic infarcts and the "ivy sign" (serpiginous sulcal hyperintensity due to slow leptomeningeal flow), present in about 80% of cases, and vessel wall imaging helps to distinguish vasculitis [18, 36]. If MRI suggests moyamoya, digital subtraction angiography (DSA) should be strongly considered, because of its greater sensitivity, because it distinguishes vasculitis and because the transdural collaterals it shows guide surgery (their preservation reduced perioperative strokes by more than 40%); complications of angiography are about 1% in high-volume centres [36]. In MSMDS, MRI is preferred to CT because radiation is a risk factor for moyamoya, and catheter angiography is reserved for focal symptoms or critical or progressive narrowing on MR angiography or Doppler [21].

The ESO suggests haemodynamic assessment (CT or MR perfusion, SPECT or PET with acetazolamide, transcranial Doppler with reactivity testing) in all patients at diagnosis, using the method most readily available at each centre; PET with oxygen-15-labelled water allows quantification of the oxygen extraction fraction and reserve, but is not widely accessible [2]. It also recommends assessing PCA involvement in all children, particularly before 5 years of age, and in all adults [2]. The Japanese guidelines recommend haemodynamic assessment only in the early postoperative period, to distinguish ischaemia from hyperperfusion [52].

5.3. Further investigation when moyamoya syndrome is suspected

The work-up includes investigation for associated diseases: examination of the skin and eyes (café-au-lait macules, Lisch nodules, iris flocculi, mydriasis), echocardiography and aortic imaging when ACTA2 is suspected, blood pressure and imaging of the renal and abdominal arteries (extracranial lesions of de novo RNF213, GUCY1A1, NF1 and Grange syndrome), full blood count and haemoglobin electrophoresis, thyroid function and autoantibodies, and, in the presence of neurological signs of interferonopathy, the type I interferon signature and CSF [1, 18, 20, 99]. The AHA/ASA recommends testing for systemic disease when there are signs: achalasia for GUCY1A1, features of trisomy 21 for karyotype, Noonan phenotype for PTPN11 and CBL, and signs of NF1 for the NF1 gene [18].

5.4. Genetic testing: indications, strategy and limitations

The guidelines diverge on testing. The ESO suggests not performing systematic screening for RNF213 p.Arg4810Lys in patients with MMA, regardless of ethnicity (expert consensus, vote 8/10), because no study has compared testing with not testing and heterozygosity has not been associated with prognosis [2]. The Japanese guidelines do not address the issue [52]. The AHA/ASA considers that testing may help to distinguish MMD from other intracranial stenoses in East Asian patients, that its role is uncertain in non-Asians and that it is of limited use for unaffected relatives, but that systemic manifestations should prompt testing for the associated disease [18]. The AHA/ASA paediatric statement notes that, in North America, less than 5% of paediatric cases have a clear genetic association, except in children of Asian ancestry, 30–50% of whom have RNF213 variants, and that identification of a variant has consequences for family screening and for surgery [36].

In practice, exome sequencing (preferably as a trio) or an extended panel with copy number variant (CNV) analysis applies when there is childhood onset, a family history, consanguinity, extracranial disease, systemic signs or East Asian ancestry [18, 20, 45, 50]. CNV analysis is indispensable for the Xq28 BRCC3 deletion [37]. In the presence of the characteristic MSMDS phenotype, targeted testing of ACTA2 codon 179 is the first step, supplemented by extended sequencing if negative [39]. RNF213 is a very large gene (the founder variant lies at position c.14429 of the coding sequence) and highly polymorphic, with rare variants also in European controls; a rare variant outside the C-terminal hotspot, transmitted by a healthy parent, is difficult to interpret [56, 58].

The limitations should be stated in the report. The diagnosis of MMD is radiological, not genetic: an RNF213 variant neither confirms nor excludes the disease, because p.Arg4810Lys is present in 0.5–2% of healthy East Asians and absent in many patients, particularly non-Asians [55]. The proportion of patients with an identified variant in a series is not equivalent to the clinical sensitivity of the test, and a negative exome does not exclude a genetic cause, because much of European and North American MMA remains without a molecular explanation [36, 45]. Parental mosaicism should be considered when a variant appears to be de novo, and deep intronic variants, repetitive regions and small structural variants escape conventional exome sequencing.

When ACTA2-related arteriopathy is suspected, an abnormally straight course of the intracranial arteries justifies ACTA2 sequencing, even without a family history of aortopathy, without other signs of MSMDS and even in the presence of a basal collateral network [94]. In carriers of ACTA2 variants outside residue 179, Kaw and colleagues propose cerebrovascular imaging when the variant is novel and de novo, when there is patent ductus arteriosus or congenital mydriasis, or when aortic surgery or dissection occurred before the age of 30 [43]. The ACMG classification of novel ACTA2 variants may remain at "uncertain significance" or "likely pathogenic" even with an overlapping phenotype. De novo origin and phenotypic concordance are criteria that weigh in reclassification, as in the case of NM_001613.4:c.137T>G p.(Met46Arg) [43]. In an infant with interstitial lung disease, mydriasis and congenital heart disease, trio exome sequencing should be requested early. In an 8-month-old infant, the result arrived only two weeks after death [42]. Prenatal diagnosis is possible when ultrasound shows megacystis or ventriculomegaly, but these findings are not specific [43].

5.5. Report content and interpretation

The report should state the gene, the versioned reference transcript (MANE Select: RNF213 NM_001256071.3; ACTA2 NM_001613.4; GUCY1A1 NM_001130682.3; ANO1 NM_018043.6 in the publications), the HGVS cDNA and protein nomenclature, the zygosity (homozygosity or compound heterozygosity confirmed by phasing for GUCY1A1 and NOS3), the classification according to the ACMG/AMP criteria and, for RNF213 p.Arg4810Lys, an indication that it is a risk allele and not a Mendelian pathogenic variant [38, 56, 57, 60, 63]. The historical numbering c.14576G>A/p.R4859K should accompany the current designation when the report refers to the Japanese literature [19, 56]. Variants of uncertain significance in RNF213 outside the RING domain benefit from segregation studies in affected and unaffected relatives, bearing in mind that familial penetrance is around 25% and that absence of disease in a carrier does not make the variant benign [58].

5.6. Practical algorithm

Sources for the algorithm: [1, 2, 18, 20, 36, 39].

5.7. Prenatal and preimplantation diagnosis

Prenatal or preimplantation genetic diagnosis is technically possible when the familial variant is Mendelian and highly penetrant (BRCC3 deletion, biallelic GUCY1A1 or NOS3 variants, ACTA2 Arg179 in a carrier parent, which is exceptional because MSMDS is almost always de novo) [9, 11, 13, 21]. For heterozygous p.Arg4810Lys, the low penetrance does not support prenatal diagnosis with useful predictive value; homozygosity, with penetrance above 78%, raises a different question that requires individual counselling [18, 19]. The sources consulted describe no prenatal diagnosis series in MMA.

6. Differential Diagnosis

6.1. Comparative table

6.2. Moyamoya disease versus atherosclerotic stenosis and vasculitis

In adults, particularly in older patients with unilateral disease, the main difficulty is to separate MMD from atherosclerotic stenosis of the terminal portion of the ICA or MCA. The 2021 criteria therefore incorporated the reduction in outer diameter on heavily T2-weighted imaging, characteristic of MMD and absent in atherosclerosis, and require angiography for unilateral lesions or those with atherosclerosis [1]. The difference has therapeutic consequences: indirect surgery forms collaterals in about 80% of adults with MMD, but not in atherosclerotic stenosis with haemodynamic failure [1]. Vasculitides and systemic autoimmune diseases cause quasi-moyamoya and should be excluded by vessel wall imaging, CSF and serology [1, 36].

6.3. Monogenic forms compared

Among the genetic forms, extracranial clues guide testing: congenital mydriasis with patent ductus arteriosus is pathognomonic of MSMDS, and its angiographic pattern (dilated proximal ICA, straight arteries, no basal network) differs from that of classic MMD [21, 53]; achalasia points to GUCY1A1, although it may be absent [11]; hypogonadism with short stature in a male points to the BRCC3 deletion [37]; and intracranial aneurysms and posterior circulation disease suggest ANO1 [38]. De novo RNF213 variants are distinguished by very early onset and by extension of the occlusions to the abdominal aorta and the renal and limb arteries, which brings them close to Grange syndrome and NF1 [20, 68].

6.4. Moyamoya syndromes in systemic diseases

In sickle cell disease, MMS is a risk marker: among children with stroke treated with chronic transfusions, those with moyamoya collaterals had more recurrent events and more recurrent strokes (21% versus 4%) [74]. In NF1, the presentation and angiographic course are comparable to those of MMD, but irradiated patients have a higher risk of stroke and perioperative complications [98]. In trisomy 21, congenital heart disease may lead to a stroke being attributed to embolism and delay recognition of moyamoya [36]. Alagille syndrome, MOPD II and SAMHD1-related Aicardi-Goutières syndrome combine stenoses and aneurysms, and in the last of these the inflammatory basis of the arteriopathy raises the possibility of immunosuppression [23, 24, 26]. Management of the underlying disease follows the specific recommendations for each disease.

Among the genetic causes of ischaemic stroke in children to consider besides moyamoya, the AHA/ASA paediatric statement highlights ACTA2 syndrome, COL4A1 variants and PHACE syndrome, and notes that mitochondrial stroke-like episodes do not respect arterial territories [36]. DADA2, described in 2014 as a monogenic autoinflammatory disease with stroke and systemic small- and medium-vessel vasculitis in young children, shares livedo reticularis with ACTA2-related disease and MYMY6, but the infarcts are lacunar and deep and there is no terminal ICA stenosis; diagnosis rests on low or absent ADA2 activity in plasma or serum, confirmed by molecular analysis [100]. The distinction has direct therapeutic consequences: in DADA2, tumour necrosis factor inhibitors (etanercept, infliximab, adalimumab, golimumab) control inflammation and reduce the occurrence of stroke, and discontinuation of acetylsalicylic acid and anticoagulants is recommended because of the risk of haemorrhagic stroke, unlike the antiplatelet therapy used in MMA [100].

7. Treatments and Risk Management Strategies

7.1. General principles

No treatment halts the progression of the steno-occlusive lesions; management seeks to reduce the risk of ischaemic and haemorrhagic stroke and disability through three instruments: surgical revascularisation, antiplatelet therapy in non-haemorrhagic disease and prevention of factors that reduce cerebral perfusion [2, 40]. No drug is approved for MMA and there is no therapy targeting any of the genes, and the surgical decision should be made by experienced multidisciplinary teams (vascular neurology, paediatric neurology, neurosurgery, neuroradiology and anaesthesia) in referral centres [2, 36, 41]. The AHA/ASA documented that high-volume paediatric centres (more than 30 cerebrovascular procedures per year) had 32% shorter hospital stays, 57% lower costs, an 8-fold higher probability of discharge home and 15-fold lower mortality [36].

7.2. General measures and prevention of triggering factors

Patients and families should know that hyperventilation, dehydration, fever and hypotension precipitate ischaemia. The Japanese guidelines advise children to avoid very hot meals, strenuous exercise, wind instruments and blowing up balloons, and in infants to avoid prolonged crying and dehydration from vomiting or diarrhoea [40]. The French PNDS (National Diagnosis and Care Protocol) adds prevention of dehydration in gastroenteritis and hot weather and control of vascular risk factors (dyslipidaemia, diabetes, hypertension, smoking), with caution in the use of antihypertensives, which may worsen hypoperfusion [41]. In MSMDS, adequate hydration is recommended and, in the event of sudden dizziness, sitting or lying down with the legs raised [21].

Raised homocysteine, body mass index and diabetes have been associated with a higher risk or ischaemic recurrence, and hypertension and dyslipidaemia with events in asymptomatic patients [52]. The Chinese consensus suggests a postoperative statin, with LDL cholesterol at or below 1.8 mmol/L, based on a non-randomised study in which atorvastatin 20 mg/day improved postoperative collaterals and on a Korean cohort with fewer ischaemic and haemorrhagic strokes on statins; neither the ESO, nor the Japanese guidelines, nor the AHA/ASA make a recommendation on statins [34, 52].

7.3. Antiplatelet therapy

The ESO suggests long-term antiplatelet therapy in non-haemorrhagic MMA to reduce the risk of embolic stroke without increasing haemorrhage (expert consensus, 9/9), acknowledging very low quality of evidence [2]. The observational studies are discordant: in a per-hemisphere study with 62 months of follow-up, infarction occurred in 2.3% with and without antiplatelet therapy and haemorrhage in 5.9% versus 7.8%; in another, with propensity matching (33 months), ischaemic stroke was less frequent on antiplatelet therapy (5.6% versus 8.4%); in the Japanese registry of 344 patients, recurrence did not differ (2.9% versus 1.6% at 5 years), but recurrent haemorrhages were more frequent without antiplatelet therapy (4.2% versus 0% at 5 years); and in the International Pediatric Stroke Study, the 20% recurrence over 13 months in 174 children did not differ with antiplatelet therapy [2]. In a Korean national database of 163,347 person-years, antiplatelet therapy was associated with lower mortality (HR 0.77; 95% CI 0.70–0.84), with greater benefit with cilostazol, but the ESO meta-analysis of 2870 patients found no effect on mortality (RR 1.07; 95% CI 0.22–5.36) [2, 18].

The Japanese guidelines accept oral antiplatelet therapy in ischaemic MMD (grade C, low evidence), with caution in prolonged use because of the risk of haemorrhagic transformation, monitoring of microbleeds on T2*, switching to clopidogrel or cilostazol if aspirin does not prevent events, and a high haemorrhagic risk with prolonged dual antiplatelet therapy, particularly with cerebral atrophy or highly developed collaterals [40]. The 2021 AHA/ASA guideline on secondary prevention considers that antiplatelet therapy, typically aspirin monotherapy, may be reasonable after ischaemic stroke or TIA in MMD (class 2b, level C-LD) [18]. The Chinese consensus gives a strong recommendation with moderate evidence for antiplatelet therapy in ischaemic MMD, weighing the haemorrhagic risk, and a conditional one for cilostazol in cognitive impairment [34]. The French PNDS indicates aspirin as first line in the absence of haemorrhage [41]. In haemorrhagic disease no guideline makes a recommendation [52]. In children, the 2012 American College of Chest Physicians (ACCP, CHEST) guideline suggests aspirin over no treatment as initial therapy for acute ischaemic stroke due to moyamoya (grade 2C) and referral to a revascularisation centre [101].

The guideline does not set an aspirin dose for adults with MMA; the Portuguese SmPC dose for prevention of TIA and cerebral thrombosis (100–300 mg/day) is the available reference, and the only dose tested in a trial in MMA is 100 mg/day, in a study not yet started [102, 103]. In children, the ACCP dose (1–5 mg/kg/day) prevails over the Portuguese label, which does not authorise use under 18 years [101, 102]. The Japanese and Korean data favouring cilostazol (better cerebral blood flow and cognition than clopidogrel in a prospective cohort, ischaemic recurrence of 6% at 5 years with cilostazol or clopidogrel, greater survival benefit) are observational, and the European label has extensive cardiovascular contraindications [18, 104]. Calcium channel blockers are not recommended by the Chinese consensus [34]. Perioperative intravenous heparin increased haemorrhage without improving bypass patency [18]; the Japanese guidelines cite low-molecular-weight heparin as an alternative in children with aspirin resistance [40].

7.4. Treatment of acute stroke

The Japanese guidelines accept intravenous thrombolysis with recombinant tissue plasminogen activator (rt-PA) in the hyperacute phase of ischaemic MMD, with careful assessment of the haemorrhagic risk (grade C, low evidence), and recommend not lowering blood pressure in the acute phase of ischaemia, glycerol for oedema with intracranial hypertension, antipyretics, anticonvulsants, glycaemic control, oxygen and normocapnia if mechanically ventilated [40]. The ESO and the AHA/ASA make no recommendation on thrombolysis or thrombectomy in MMA, owing to lack of data [52]. In acute haemorrhage, the Japanese guidelines accept lowering systolic pressure as in spontaneous intracerebral haemorrhage, weighing concomitant ischaemia (grade C), and stopping antiplatelet agents or anticoagulants, with ventricular drainage when indicated [40]. In a child with stroke due to moyamoya, hypotension should be avoided and actively treated [36]. In sickle cell disease, acute treatment includes hydration and correction of hypoxaemia and hypotension [36].

7.5. Surgical revascularisation: indications

Revascularisation is the primary treatment of symptomatic MMD and MMS [36]. The direct technique anastomoses a branch of the superficial temporal artery (STA) to a cortical branch of the MCA (or ACA) and immediately increases flow; indirect techniques place vascularised tissue (dura mater, temporalis muscle, galea, the STA itself, omentum) in contact with the cortex to stimulate the formation of new vessels, with an effect over weeks to months (3–6 months even in children); combined techniques bring the two together [2, 18]. Indirect techniques include encephalomyosynangiosis (EMS), encephaloarteriosynangiosis (EAS), encephaloduroarteriosynangiosis (EDAS), encephaloduroarteriomyosynangiosis (EDAMS), pial synangiosis and multiple burr holes [2, 40].

In adult haemorrhagic MMA, the ESO recommends revascularisation (with evidence only for direct STA–MCA bypass) when there is haemodynamic compromise and choroidal collaterals (low-quality evidence; weak recommendation in favour) [2]. The Japanese guidelines consider it reasonable, particularly in posterior haemorrhage (grade B, moderate evidence), and the 2023 AHA/ASA statement cites class 2a [40, 52]. In adult ischaemic MMA, the ESO maintains uncertainty in the evidence and suggests, by consensus, revascularisation when there are symptoms and/or markers of haemodynamic compromise; the Japanese guidelines consider it reasonable (grade B, low evidence); the 2021 AHA/ASA guideline considers that direct or indirect revascularisation may be beneficial to prevent ischaemic stroke or TIA (class 2a, level C-LD, according to the 2023 statement) [2, 18, 40]. In asymptomatic adults, the ESO suggests conservative management, except when haemodynamic compromise and silent ischaemic lesions coexist in the same region [2].

In children, the ESO suggests revascularisation when there are persistent ischaemic symptoms or haemodynamic compromise, and early surgery for recurrent TIAs or ischaemic strokes, except with a large territorial lesion [2]. The AHA/ASA paediatric statement indicates surgery in the presence of stroke, TIA or other evidence of compromised flow or reserve, with relative contraindications in very early arteriopathy with normal perfusion and in profound medical or neurological compromise [36]. The Chinese consensus encourages early surgery in children to protect neurodevelopment and includes cognitive impairment, involuntary movements, headaches and seizures among the indications [34].

7.6. Revascularisation: technique, timing and outcomes

Regarding technique, the ESO suggests direct or combined rather than indirect revascularisation in adults, and combined revascularisation in children whenever technically possible, to reduce the short-term risk of stroke [2]. The Japanese guidelines consider indirect surgery alone insufficient in adults [40]. The ESO meta-analysis of five studies in adults found a non-significant reduction in stroke with the direct or combined technique (RR 0.58; 95% CI 0.24–1.39), with very low certainty; one of the studies favoured the indirect technique (24% versus 12%) [2]. In children, the superiority of the combined technique is not evident (0% stroke in both groups in a series of 64 children followed for 6.6 years; 5% versus 2% in another of 108), and a 2017 decision analysis attributed more quality-adjusted life years at 10 years to the indirect technique [2, 36]. Perioperative complications do not differ between techniques [2].

Regarding timing, the ESO suggests waiting 6–12 weeks after an acute event, meanwhile avoiding dehydration, fever, hyperventilation and hypotension, and accepts earlier surgery in children with recurrent TIAs or stroke with rapid and complete recovery; in a retrospective study, complications were 39.3% with surgery before 90 days versus 13.8% afterwards, but new stroke during the wait was 31% versus 11% (p=0.06) [2]. The Chinese consensus postpones surgery for 1–3 months after a recent ischaemic event and performs revascularisation 1–3 months after resorption of a haemorrhage, preserving the STA if haematoma evacuation is needed [34]. The AHA/ASA paediatric statement advises operating as soon as possible, accepting delays of weeks to organise the team or allow oedema to resolve [36].

The outcomes support benefit. In a retrospective multicentre study of 282 children, surgery was associated with fewer strokes (2.8% versus 13.2%; RR 0.21; 95% CI 0.08–0.57) and less disability (2.8% versus 20.6%; RR 0.14; 95% CI 0.05–0.34) over 41 months [2]. In a follow-up of 93 patients 10.5 ± 4.4 years after combined surgery, late haemorrhage was 0.10% per patient-year and contralateral or PCA progression 1.5% per patient-year [2]. In NF1, 95% of the 22 children followed for more than 6 months after pial synangiosis (mean 80.2 months) remained stable or improved, despite angiographic progression in 48% [98]. Endovascular revascularisation is not advised: angioplasty and stenting had 25% success and devastating haemorrhage in 7%, and the Japanese guidelines and the AHA/ASA advise against them [18, 52].

7.7. Perioperative care and anaesthesia

Anaesthesia in a patient with MMA, even for minor procedures, should be planned with the referral team [35, 41]. The AHA/ASA recommends intravenous hydration the day before, an arterial line before induction, systolic pressure at or above the asymptomatic baseline (avoiding more than 180 mmHg), normocapnia with end-tidal CO2 between 35 and 45 mmHg, euvolaemia or mild hypervolaemia and avoidance of mannitol [18]. In children, hydration at 1.25–1.5 times maintenance and prevention of hyperventilation caused by crying, pain or vomiting are central [36].

Cerebral hyperperfusion syndrome occurred in 16.5% (95% CI 11.3–22.3%) of 2225 patients from 27 studies (3.8% in children, 19.9% in adults), with transient deficits (70.2%), haemorrhage (15.0%) and seizures (5.3%); it is controlled with systolic pressure below 130 mmHg, and perioperative minocycline or edaravone may prevent it [18, 40]. The Chinese consensus records transient postoperative neurological dysfunction in 16.5–38.2% and advises perioperative antiseizure medication [34]. The ESO considers continuation of aspirin monotherapy during bypass safe; if stopped, it should be restarted 1–7 days after surgery depending on CT, and with dual antiplatelet therapy clopidogrel (or the second antiplatelet agent) is stopped 7 days before [2]. Postoperative aspirin was associated with greater bypass patency in one study (98.7% versus 89.7%) but not in another [2].

In MYMY6, an episode of microvascular dysfunction with cyanosis after sevoflurane anaesthesia and the recommendation to screen for the angiopathy before myotomy for achalasia underline the perioperative risk [35]. In MYMY8, a stroke occurred during an orchidopexy [13].

7.8. Management by genetic entity

MSMDS (ACTA2 Arg179). The recommendations of Regalado and colleagues call for coordinated multidisciplinary care, because drugs used for one organ may worsen another [21]. On the neurological side, they recommend baseline brain MRI and MR angiography of the head and neck, periodic cerebrovascular screening (at an interval not yet established), transcranial Doppler as a possible surveillance tool and catheter angiography only in the presence of focal symptoms or critical or progressive narrowing; aspirin may be considered for primary prevention in asymptomatic patients with vasculopathy, with monitoring for Reye syndrome; anticoagulation in carriers of a mechanical valve prosthesis or with thromboembolism; close monitoring during febrile infections and hypoxaemia; and routine cognitive and psychological assessment [21]. Cerebral revascularisation in MSMDS was followed by more postoperative ischaemic strokes than in MMD; three patients with indirect surgery formed collaterals and improved perfusion without an overall reduction in the risk of stroke or TIA, and one died of stroke three years later [21]. The AHA/ASA paediatric statement describes this population as very high risk, to whom indirect surgery is offered after discussion with the family and with strict intraoperative blood pressure control [36]. On the cardiovascular side, recommendations include surgical closure of the ductus arteriosus in infancy (delay worsens pulmonary hypertension), assessment of the entire aorta at diagnosis, surveillance of the descending and abdominal aorta from 10 years of age (or earlier if there are abnormalities), postponement of root surgery until the annulus can accept an adult graft, discussion of valve-sparing surgery at 45 mm and annual ultrasound surveillance of peripheral aneurysms, repairing those that are symptomatic, thrombosed or larger than 20 mm; no drug has been tested specifically in ACTA2 carriers [21, 39]. In planning aortic repair in ACTA2 carriers, Clinical Cardiogenetics recommends allowing for the more extensive involvement of the root, ascending aorta and arch and considering repair of all these regions, even if one of them is not yet dilated; in HTAD in general, it recommends drugs that reduce haemodynamic stress, such as beta-adrenergic blockers, and more severe phenotypes are noted with residue 179 substitutions and p.Arg258 variants [39].

MSMDS: revascularisation and medical alternatives after 2018. In the Zurich study, 8 consecutive children with neurological instability and confirmed haemodynamic compromise underwent indirect or combined direct and indirect surgery, and 2 stable children continued conservative treatment. Postoperative revascularisation reached Matsushima grade A or B in 6 patients, with patent and sufficient surgical collaterals up to 14 years later [93]. The authors conclude that surgical neovascularisation is comparable to that in idiopathic MMD and may support revascularisation in symptomatic children with haemodynamic compromise. The study is small, without a comparison group, and only the abstract has been published so far [93]. In a 46-year-old woman (Tokyo) with p.(Arg179His) and an infarct of the right cingulate gyrus, a triple direct bypass (three anastomoses of the right superficial temporal artery to the middle and anterior cerebral arteries) was complicated by postoperative ischaemic stroke. The authors attribute the difficulty to the fragility of the donor artery and to poor collateral development, but retain direct bypass as an option in the presence of multiple progressive strokes [105]. In an 11-year-old boy with repeated TIAs, proximal ICAs with fusiform dilatation, terminal segmental stenosis, straightened distal branches and increasing white matter lesions, the team opted for bosentan, an oral endothelin receptor antagonist, instead of surgery, because of the perioperative risk. The ischaemic episodes ceased and the white matter lesions did not increase over 7 years [106]. It is an isolated case, without a control. The abstract does not state the dose. In patients with MSMDS and pulmonary hypertension, the management described is symptomatic: diuretics, sildenafil or bosentan, closure of the ductus arteriosus and repair of coarctation, oxygen and treatment of infections in interstitial lung disease [42].

Bosentan: reference dose and monitoring. There is no published bosentan dose for ACTA2-related arteriopathy. The reference is the European SmPC of Tracleer® (bosentan) in pulmonary arterial hypertension (PAH), an extrapolation from another condition treated with the same drug [107]. In adults, the SmPC indicates 62.5 mg twice daily for 4 weeks and then 125 mg twice daily as maintenance. In children aged 1 year or older, the initial and maintenance dose is 2 mg/kg morning and evening. Paediatric concentrations did not rise with doses above 2 mg/kg or with three daily doses, and the SmPC does not establish a dose for persistent pulmonary hypertension of the newborn. The SmPC defines no maximum dose beyond maintenance of 125 mg twice daily in adults and 2 mg/kg twice daily in children [107]. Aminotransferases are measured before starting treatment, then every month and 2 weeks after each dose increase. Haemoglobin is measured before starting, every month for the first 4 months and every 3 months thereafter [107]. Bosentan is contraindicated in pregnancy, in moderate to severe hepatic impairment, with baseline aminotransferases above 3 times the upper limit of normal and with ciclosporin A. In women of childbearing potential it requires reliable contraception and monthly pregnancy tests, and it reduces the efficacy of hormonal contraceptives [107]. In MSMDS, hypotension and dysautonomia (section 3.6) add risk to any systemic vasodilator, and the decision should be made by a team experienced in the syndrome.

ACTA2-related disease without Arg179. Aortic management follows gene-specific thresholds (root or ascending aortic surgery at ≥4.5 cm without risk factors and at ≥4.2 cm with risk factors, according to the 2022 ACC/AHA guideline cited by GeneReviews and GeneGuidance), and in carriers of variants that predispose to occlusive disease screening for coronary and cerebrovascular disease is recommended [22, 59]. In patients with the SMDS-like spectrum (ductus arteriosus or congenital mydriasis without the full picture), Kaw and colleagues recommend the same vascular surveillance as in MSMDS, annual ophthalmological examination and consideration of simultaneous repair of a still-normal arch when the root and ascending aorta are repaired [43]. In the family with p.(Arg198Cys), the proband with hypoperfusion in the right MCA territory underwent superficial temporal artery to MCA bypass and aspirin 100 mg/day [94]. One-third of patients with ACTA2 had dissection at diameters below 5.0 cm, which supports repair at 4.5 cm [94].

MYMY6 (GUCY1A1). The authors of the initial description recommend screening GUCY1A1 in isolated early-onset achalasia and, if a pathogenic variant is found, performing 3D TOF MR angiography to detect asymptomatic angiopathy, particularly before surgery for achalasia [35]. Hypertension was treated with nicardipine and perindopril in one patient [35]. The loss of the α1β1 isoform with preservation of the α2β1 isoform in some tissues led the authors to propose phosphodiesterase 5 inhibitors (sildenafil) for Raynaud phenomenon and erectile dysfunction, NO donors or sGC stimulators and activators; these proposals have not been tested [35].

MYMY4 (BRCC3). Management includes, besides the angiopathy, surveillance of the dilated cardiomyopathy, growth hormone deficiency, hypogonadism and cataracts described in the patients [9, 37].

Moyamoya syndromes in systemic diseases. In sickle cell disease, regular transfusions (haemoglobin S below 30%) or hydroxycarbamide after one year of transfusions are the primary prevention of stroke when there is no MMS on MR angiography; medical therapy is ineffective in sickle cell MMS and surgery is an option [36, 40]. In NF1, pial synangiosis has proved safe and effective, and irradiated patients have a higher perioperative risk [98]. In MOPD II, moyamoya was treated by EDAS or pial synangiosis, mainly in childhood, and the aneurysms by clipping, embolisation or stenting [23]. In Alagille syndrome, revascularisation has been described with good results [108]. In SAMHD1-related Aicardi-Goutières syndrome, the occlusive arteriopathies can be revascularised and the aneurysms clipped or embolised, and the inflammatory basis raises the possibility of immunosuppression [24].

7.9. Agents and situations to avoid

Drugs that induce hypotension or vasoconstriction should be avoided: nasal vasoconstrictors, triptans and ergot derivatives, by any route [41]. The AHA/ASA notes that headaches, which are frequent and often migraine-like, should not be treated with drugs that cause vasoconstriction, inhibit vasodilatation or lower blood pressure; the Chinese consensus considers non-steroidal anti-inflammatory drugs a reasonable choice [18, 34]. Antihypertensives are used with caution [41]. Unnecessary irradiation should be avoided, particularly in NF1, in which radiotherapy is used only when absolutely necessary (also because of the risk of malignant peripheral nerve sheath tumour), and in MSMDS MRI is preferred to CT [21, 51, 109]. Risk situations include dehydration, fever, hyperventilation (crying, anxiety, wind instruments), hypotension and unplanned anaesthesia [2, 40, 41].

7.10. Pregnancy and breastfeeding

The Japanese guidelines consider pregnancy and delivery acceptable in MMD, with close cooperation between obstetrics, paediatrics and neurology, because an increased risk of stroke in pregnancy, delivery and the puerperium cannot be excluded [40]. In a Japanese national survey, perinatal neurological events occurred in 5.1% of 59 deliveries in women with known MMD (morbidity 1.7%), and in a patient survey in 6.6% of 76 deliveries with a prior diagnosis and 2.0% of 202 deliveries without a diagnosis, without permanent morbidity; caesarean section was used in about 70–76%, with no difference in events between caesarean and vaginal delivery; among 5 women diagnosed after a perinatal event, 3 were left with disability and one died of haemorrhage [110]. In a systematic review of 54 publications, 95.2% of women with known MMD had good outcomes and there were no poor fetal outcomes; when MMD was diagnosed because of an event in pregnancy (on average at 28.7 weeks, haemorrhagic in 69.5%), maternal mortality was 13.6% and fetal death 23.5%; the authors concluded that MMD does not contraindicate pregnancy if blood pressure and ventilation are controlled and that there is no evidence of benefit from prior bypass or from caesarean section [111]. In MSMDS, pregnancy is high risk (aortic dissection, stroke, atonic uterus with insufficient contraction for vaginal delivery), and cardiological and high-risk obstetric assessment before conception is recommended [21]. In ACTA2-related disease, 20% of aortic dissections were pregnancy-related in one series, and preconception repair at 4.5 cm is recommended [59]. Cilostazol is contraindicated in pregnancy and aspirin at doses above 100 mg/day in the third trimester [102, 104]. The sources consulted give no specific guidance on breastfeeding in MMA.

7.11. Transition to adulthood

Paediatric disease continues to progress in adulthood, with progression to the PCA up to 15 years after surgery, and imaging follow-up should be maintained in the long term [2]. In MSMDS, genetic counselling of adolescents and young adults on inheritance, recurrence and reproductive options is part of care, and mortality rises after the age of 30 [21]. The neurodevelopmental disorder genes (CHD4, CNOT3, SETD5) show that cerebrovascular complications may arise in adulthood in patients followed in childhood for another reason [45].

7.12. Regulatory status

There are no medicines authorised by the FDA, the European Commission or INFARMED for MMA or for any monogenic form in the sources consulted; the antiplatelet agents used are medicines authorised for other cardiovascular indications.

7.13. Available intervention evidence

The only randomised trial of surgery is JAM, limited to Japanese adults with haemorrhage and mRS 0–2, with the sample reduced from 160 to 80 patients and a primary endpoint of marginal significance (p=0.057); its generalisability to European patients, to children and to ischaemic disease has not been demonstrated [2, 52]. There are no randomised trials of revascularisation in ischaemic disease, which the AHA/ASA and others consider ethically difficult to perform [52]. The relative reductions should be read together with the absolute rates (in JAM, 5 percentage points per year of difference in the composite endpoint), and it is not legitimate to compare indirectly trials and cohorts with different populations and endpoints. The remote ischaemic conditioning trial is a small single-centre pilot study, without a haemodynamic difference. The Chinese trials of surgery in haemorrhagic MMD (NCT02319980, phase 3, 360 patients planned) and of dl-3-n-butylphthalide (NCT04205578) have unknown status on ClinicalTrials.gov, with no results [115].

7.14. Developments in the last 6–12 months

Preclinical (2025). In cultured Acta2 R179C/+ SMCs, nicotinamide riboside, a precursor of nicotinamide adenine dinucleotide (NAD+), increased basal, ATP-linked and maximal oxygen consumption, reduced lactate production and glucose flux through glycolysis, the pentose phosphate pathway and the Krebs cycle, restored the quantity and activity of respiratory chain complexes I and IV, increased contractile markers and decreased migration, without altering proliferation or increasing apoptosis or necrosis; the effect was mediated by increased complex function and not by mitochondrial biogenesis, because mitochondrial DNA and mitochondrial labelling did not rise [61]. In mice with Acta2 R179C/+ in SMCs treated with 1000 mg/kg on alternate days, from 5 days before left carotid ligation until the 21st postoperative day, there were no deaths (0 of 16, versus 8 of 28 untreated females; p=0.02), day 1 neurological deficit scale scores fell, neointimal lesions and intraluminal P-selectin-positive cells decreased, loss of hippocampal and striatal neurons was prevented and remodelling of the leptomeningeal collaterals was attenuated to values close to those of controls; the narrowed geometry of the circle of Willis and the 20–25% reduction in ipsilateral cerebral blood flow were not reversed [61]. The limitations stated by the authors are the dose used, because previous pharmacokinetic studies in mice showed similar efficacy between 100 and 1000 mg/kg and off-target effects cannot be excluded, the absence of cognitive or memory testing to confirm the functional deficit corresponding to the hippocampal neuronal loss, the impossibility of excluding effects of nicotinamide riboside on other cell types and the differences between mouse and human cerebrovascular anatomy [61]. The appearance of occlusive lesions with only 67% mutant SMCs suggests that a genome editing strategy would prevent cerebrovascular disease only if it corrected almost all mutant SMCs, because the cells that migrate into the neointima tend to be the mutant ones [61].

Gene editing in MSMDS: scale and rationale of the approach. The most frequent MSMDS variant, NM_001613.4:c.536G>A p.(Arg179His), replaces a C•G pair with a T•A pair and is therefore correctable by an adenine base editor, which converts A•T to G•C through deamination of a single adenine, without a double-strand DNA break and without the need for homologous recombination repair; the two teams that developed the strategy chose this class of editing and not prime editing, the conventional CRISPR-Cas9 nuclease, allele-specific silencing by antisense oligonucleotides or RNA interference, or AAV gene transfer of a normal copy of the gene [116, 117]. The rationale derives from the dominant negative mechanism: the mutant protein is incorporated into the filaments and destabilises the actin cytoskeleton of SMCs, pericytes and the other cells that express smooth muscle α-actin, so that adding normal copies of the gene does not cancel the effect of the mutant allele and the correction must act on the pathogenic base itself; precise correction also avoids the risks of transgene overexpression [117]. Prime editing and the Cas9 nuclease were used only to build the cell and animal models, not as therapy [116, 117]. None of the sources consulted describes a programme of allele-specific silencing, RNA interference or gene addition for ACTA2, and none describes gene editing or gene therapy targeting RNF213, GUCY1A1, ANO1 or BRCC3.

In vivo correction with AAV9 and a smooth muscle-specific promoter (2025). In the first of the two strategies, the ABE8e editor coupled to an SpCas9 variant with altered PAM specificity (VRQR) corrected the variant in isogenic human induced pluripotent stem cell lines with an on-target efficiency of 81.67% and 2.50% bystander editing, with off-target editing of 0.1% or less at the eight candidate sites tested and a normal karyotype after editing; SMCs derived from these corrected lines normalised proliferation, migration, contraction in collagen gel, the ratio of filamentous to globular actin, the transcriptional profile and the expression of matrix metalloproteinase 9, which was increased 2.9-fold in R179H/+ cells [116]. For the in vivo phase, the authors generated a humanised mouse with the human sequence c.536G>A in exon 6 of murine Acta2, which reproduced the multisystem disease: low systolic and diastolic blood pressure at 7 weeks, dilatation of the ascending aorta, medial tears in about 30% (5 of 16) at 6 months, dilated bladder in about 80%, dilated small intestine with disorganised villi in about 70% and hydronephrosis in about 40% of animals [116]. The editor, at 5.1 kb, exceeds the packaging capacity of an AAV (about 4.7 kb) and was therefore split into two AAV9 vectors reassembled within the cell by trans-splicing inteins, under a chimeric SM22/telokin promoter selected from four candidates for directing expression to vascular and visceral smooth muscle with minimal expression in liver and heart; administration was a single intravenous dose, via the facial vein, on the 3rd day of life, totalling 9×10¹³ vg/kg [116].

Magnitude of effect and limitations of the first strategy. In treated mice, blood pressure normalised (systolic 129.0±3.7 mmHg versus 96.7±3.7 in untreated animals and 132.9±6.1 in normal controls; diastolic 88.1±4.4 versus 57.1±2.9 and 87.3±6.8), the diameter of the ascending aorta fell from 1.49±0.05 mm in untreated animals to 1.33±0.03 mm, identical to the 1.32±0.03 mm of controls, and medial thickness recovered from 48.5±1.0 µm to 62.1±1.6 µm, close to the 64.8±2.7 µm of controls, with normalisation of collagen deposition and elastin fragmentation [116]. The morphology of the bladder, intestine and kidney also approached normal, and treated animals resisted an angiotensin II infusion of 1 µg/kg per minute for 14 days that dilated the aorta of untreated animals; there was no rise in aminotransferases, urea or creatinine [116]. Editing efficiency in the tissues was, however, low: 2.10±0.74% in the aorta (maximum about 5%), 1.11±0.29% in the bladder, 1.45±0.91% in the intestine and 1.16±0.50% in the kidney, well below the 10–30% obtained by the same group in the myocardium, which the authors attribute to SMCs representing only 10–20% of aortic tissue and to the hypothesis that mosaic restoration of the contractile phenotype is enough to slow progression [116]. The limitations stated are the milder aortopathy of the model compared with the human disease, possibly because of reduced penetrance in the hybrid genetic background, the anatomical differences between mouse and human, the possibility of AAV-related adverse effects, the unexplained mechanism of the benefit disproportionate to the editing achieved and the fact that the design assesses prevention, with treatment on the 3rd day of life, without demonstrating that the correction prevents aneurysms throughout life [116].

Editor designed for the mutant allele and capsid with smooth muscle tropism (2026). The second strategy started from a safety problem: with wild-type SpCas9, correction was accompanied by frequent bystander editing that introduced the substitution p.Met178Val, which, tested in primary human SMCs, raised the expression of the histone deacetylase HDAC9 and disorganised the actin filaments just as p.Arg179His itself did, that is, it turned the correction into a counterproductive outcome; in addition, an off-target site located in an exon of ACTC1, which encodes cardiac α-actin, was edited with an efficiency of about 10% and produced the same Met178Val substitution in the cardiac protein [117]. The authors carried out protein engineering on the mutant sequence and obtained an SpCas9-VRQR variant with the S55R substitution (eVRQR) that shifts the editing window, maintains PAM selectivity and maximises correction: about 45% on-target correction in the homozygous cell model, about 25-fold less p.Met178Val bystander editing and off-target editing below 3% at 5 of 117 validated sites, all in intergenic regions, in long non-coding RNAs or in an intron of KDM4C, with the rest at the limit of detection [117]. Nomination of the off-target sites combined in silico prediction, an unbiased cellular assay and a biochemical assay specific for base editors applied to genomic DNA from three fibroblast lines of patients with MSMDS, and more than 80% of the reads fell at the target site, edited about 8.5 times more efficiently than any off-target site [117].

Survival and phenotypic reversal in the conditional model. The model used was a mouse with the Acta2 R179H allele activated by Cre under the Myh11 promoter, restricted to SMCs, with a median survival of 6 weeks from toxic megacolon, insufficient weight gain, poorer performance on the rotarod and in the open field, a wider aorta, a narrower distal internal carotid artery, mydriasis, hydronephrosis, distended bladder and intestine and hypomyelination of the cerebral white matter [117]. The editor, split into two AAV vectors and reassembled by inteins, was administered intravenously on the 3rd day of life, 8×10¹⁰ vg of each vector, comparing the AAV9 capsid with the AAV-PR capsid, designed to transduce pericytes and SMCs; correction was up to about 60% in the liver, 22% in the heart, 23% in the cerebral vasculature and below 3% in the aorta, with AAV-PR outperforming AAV9 in all tissues and about 4-fold in the cerebral vasculature, despite fewer viral copies in the tissues, indicating greater efficiency per vector genome [117]. At the prespecified time point of 8 weeks, all treated animals and none of the untreated animals were alive, and on extended follow-up median survival rose from 6 to 22.6 weeks, with animals alive up to 44 weeks, an about fourfold gain; the deaths among treated animals resulted from intestinal and bladder distension requiring euthanasia, with no aortic dissections or strokes recorded [117]. Treatment restored weight, locomotor performance, aortic diameter on echocardiography, elastin tortuosity of the anterior and posterior cerebral arteries, the morphology of the thalamic and hippocampal arterioles, cortical myelination, the morphology of kidney, lung and liver, filamentous actin content and smooth muscle α-actin expression, and suppressed HDAC9 in the aorta [117]. Injection on the 14th day of life gave less editing and only partial improvement in survival, weight, physical performance and aortic diameter, which defines a gradient of benefit according to age at treatment [117]. The limitations stated are the difficulty of assessing the biological consequences of bystander and off-target editing in transformed cell lines and in a murine genome, the need for assays in primary human cells or organoids and for prolonged patient monitoring, the need to split the editor into two vectors, and the risk of genomic toxicity and adaptive immune response arising from sustained expression of the editor from the AAV episome, which the authors propose to address with temporal inactivation of expression or with transient delivery by messenger RNA nanoparticles or virus-like particles [117].

Clinical status. There is no human trial of gene editing or gene therapy in MSMDS, moyamoya disease or RNF213-associated angiopathy: a search of ClinicalTrials.gov identifies no interventional studies with these technologies in these conditions, and the only studies registered specifically for MSMDS are the observational natural history study of genetic vasculopathies NCT06552052 (Massachusetts General Hospital, recruiting since May 2024, 50 participants planned), the phase 1 nicotinamide riboside trial NCT06280482 and the single-patient sapropterin study NCT07574034, not yet started [113, 115, 118, 119]. Both base editing strategies are in the preclinical phase, in academic institutions and without a declared industrial sponsor for a clinical programme; a negative search does not prove non-existence, but none of the sources consulted announces a regulatory submission date.

Efficacy, risks and practical implications. The efficacy demonstrated is prevention, in animal models treated in the first days of life, of aortopathy, cerebral arteriopathy, demyelination and early death, with partial reversal when treatment is later; none of the studies demonstrated reversal of established arterial stenoses, of fixed neurological deficits or of severe intestinal and bladder dysfunction, which continued to determine the mortality of treated animals [116, 117]. The risks identified are of three kinds. Bystander editing may create a second pathogenic allele, and off-target editing in an exon of ACTC1 may affect cardiac α-actin, which shows that safety depends on the enzyme and the positioning of the editing window and not on the technology alone; with the enzyme designed for the mutant allele these events fell below 3% and in regions with no known consequence, but they were assessed only in transformed cells and in a murine genome [117]. Immunogenicity is twofold, against the AAV capsid and against the editor protein expressed persistently from the episome, with an additional risk of continued editing; there is also exposure to a high vector dose, not transposable to humans from the murine values [116, 117]. Vascular delivery is the main obstacle: SMCs are poorly transduced and correction in the aorta did not exceed 3–5% even with a specific promoter or modified capsid, and the phenotypic benefit disproportionate to this minimal editing has no established explanation, so it cannot be assumed to hold in humans [116, 117]. This observation diverged from the prediction made from the R179C/+ model, according to which only correction of almost all mutant SMCs would prevent cerebrovascular disease, because it is the mutant cells that migrate into the neointima; the discrepancy remains experimentally unresolved [61, 116, 117]. In clinical practice and counselling, the approach should not be presented as an available therapeutic option or with a foreseeable timeline, no compassionate use is possible, and management remains imaging surveillance, revascularisation in patients with haemodynamic compromise and management of the aortopathy; it is useful to direct families to natural history studies and genetic vasculopathy registries, because longitudinal characterisation is a prerequisite for a future trial. The therapeutic window suggested by the animal data precedes irreversible injury, which increases the value of early diagnosis — possible through congenital mydriasis combined with exome sequencing — and justifies targeted testing in newborns from families with a known variant [116, 117]. Counselling should make explicit that the editor was designed for the c.536G>A transition and does not serve the p.(Arg179Cys) or p.(Arg179Ser) variants, which result from different nucleotide changes, and that reproductive decisions, including prenatal and preimplantation diagnosis, should not rest on the expectation of gene therapy in the short term [117]. The authors themselves position MSMDS as a model for a broader class, including CADASIL, amyloid angiopathy, moyamoya disease, Marfan syndrome, Loeys-Dietz syndrome and vascular Ehlers-Danlos syndrome, for none of which a validated editing target exists [117].

Ongoing clinical trial. The phase 1 trial NCT06280482 (University of Texas, Houston; recruiting since March 2024; 15 patients aged 1 year or older with SMDS; estimated primary completion in July 2027) tests oral nicotinamide riboside at a weight-adjusted dose (value not stated in the registry), once daily with food for 60 days, with endpoints of blood pressure, cerebral perfusion by near-infrared spectroscopy during a tilt test, cognition, autonomic symptoms, headache, lung function and aortic diameter; no results have been published [113].

N-of-1 report (2026). On the basis of molecular simulation and in silico screening, sapropterin (Kuvan®, a tetrahydrobiopterin analogue approved for phenylketonuria) was proposed as a stabiliser of R179H actin and improved actin organisation in patient fibroblasts; in a 6-year-old girl with MSMDS and progressive cerebrovascular disease, off-label sapropterin (10 mg/kg/day, increased to 20 mg/kg/day after 2 weeks) was associated over 18 months with lower transcranial Doppler velocities, greater vessel calibre on MR angiography, an improved ratio of the supraclinoid to the petrous ICA and better physical quality of life, with no adverse effects; this is a single patient, without a control, and does not support clinical use outside research [112].

Clinical imaging study (October 2026). The longitudinal angiographic study of 10 children with ACTA2-related arteriopathy (Zurich) described moyamoya-type collaterals in 5 patients at follow-up and Matsushima grade A or B surgical revascularisation in 6 of 8 operated patients, patent for up to 14 years. The article was published online ahead of print and, for now, only the abstract is available [93].

Preclinical (2026), digestive tract. In the ACTA2 R179H mouse, which reproduces MSMDS, there was dilatation of the caecum and colon, a shorter intestine, altered colonic migrating motor complexes, slower transit and a reduced contractile response to electrical and pharmacological stimuli. The transcriptome showed changes in actin cytoskeleton genes in SMCs, and there was greater lymphocytic infiltration, without evident changes in the enteric nervous system [89].

Registered trial, not yet started. The randomised trial NCT07762547 (Beijing Tiantan Hospital; planned start in July 2026; 724 adults with mild MMD or MMS) compares aspirin 100 mg/day with placebo and remote ischaemic conditioning with a sham procedure over one year [103].

The ClinicalTrials.gov search (67 studies) found no drug trials targeting RNF213, GUCY1A1 or BRCC3, and no drug suspensions or withdrawals in MMA [115]; a negative search does not prove non-existence.

7.15. Summary of recommendations and divergences

Sources for the table: [2, 18, 34, 36, 40, 41, 52].

8. Reproductive Options and Family Counselling

8.1. Recurrence risks by scenario

Counselling depends on whether a genetic cause has been identified and on its nature. In MMD without an identified cause, the risk to relatives is increased but cannot be quantified precisely: 10–12% of Japanese patients have a family history, about 7–12% in Western cohorts, and familial clustering is rare in white patients (3.4% in a large North American series) [30, 36, 38]. In clustering studies, the risk is highest for the twin of an affected individual, followed by a sibling, the mother and, lastly, the father [18].

8.2. Screening of relatives and cascade testing

The AHA/ASA paediatric statement reserves MRI with MR angiography of unaffected relatives for first-degree relatives of patients who have other first- or second-degree relatives with moyamoya, for those with a suggestive history (TIA, stroke, severe headaches or unexplained seizures) and for identical twins; without symptoms, it is reasonable to postpone imaging until the child tolerates MRI without anaesthesia, and the interval after a normal MRI is unknown, because normal scans may later show symptomatic disease [36]. The French PNDS allows imaging of asymptomatic relatives only after a genetic counselling and vascular neurology consultation [41]. When a Mendelian variant is identified (ACTA2, GUCY1A1, BRCC3, NOS3, ANO1), cascade testing allows surveillance to be concentrated on carriers; for RNF213 p.Arg4810Lys, the AHA/ASA considers the benefit limited in unaffected relatives, given the incomplete penetrance [18]. The AHA/ASA paediatric statement cites, for carriers of an RNF213 variant in families with moyamoya, a probability close to 50% of manifesting arteriopathy and a familial penetrance of about 23%, figures that show the variability of estimates depending on study design [36].

In families with ACTA2, screening of relatives includes the aorta and, for variants with occlusive disease, the coronary and cerebral circulations [22, 59]. In the German family with p.(Arg198Cys), cascade testing identified the variant in the mother (61 years), in a daughter (12 years) and in a son (9 years) of the proband; the son's dizygotic twin and a maternal aunt were not carriers [94]. The three carriers were asymptomatic, but all had straight intracranial arteries on MRI. The mother also had bilateral stenoses of the terminal ICA and the proximal MCA and ACA and white matter hyperintensities, with chest CT showing no aortic dilatation. She subsequently had a dissection of the descending thoracic aorta, treated with a stent, and died in the same year from aortic rupture [94]. The family shows that cerebral and aortic imaging of carriers identified by cascade testing has clinical consequences, and that an aorta without dilatation does not exclude dissection. In the family with NM_001613.4:c.115C>T p.(Arg39Cys), the same inherited variant caused aortopathy with ductus arteriosus in several relatives and, in the proband, pulmonary, digestive and urinary disease without vascular disease. The authors allow for a second unidentified variant in another gene [43]. In families with early achalasia, a sibling with achalasia should be tested for GUCY1A1 and investigated with MR angiography [35].

8.3. Carriers, mosaicism and minors

Heterozygotes for GUCY1A1 and NOS3 have no described cerebrovascular disease in the published families [13, 35]. Female carriers of the Xq28 deletion have not been described as affected [37]. The possibility of germline mosaicism should be mentioned when the variant is de novo (ACTA2 Arg179, de novo RNF213, DIAPH1, CHD4, CNOT3). Predictive testing in minors is justified when it changes surveillance: in MSMDS and in BRCC3 deletion the diagnosis is clinical from an early age; in families with a known Mendelian variant, the result guides cerebral and aortic imaging; for isolated p.Arg4810Lys, the predictive value is low and the decision should be individual [18, 36]. In the heritable aortopathies, Clinical Cardiogenetics accepts predictive testing of minors in childhood-onset diseases with available preventive measures, considers the benefit less evident in non-syndromic forms, in which testing is often deferred to adulthood to protect future autonomy, and always calls for psychosocial support and pre-test counselling on risks, benefits and limitations; in ACTA2-related disease, in which stroke and moyamoya may appear in childhood, this judgement is made case by case [39].

8.4. Prenatal and preimplantation diagnosis

Prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic testing are possible in the Mendelian forms with a known familial variant [9, 11, 13]. In RNF213 susceptibility, the low penetrance in the heterozygous state does not support these options; homozygosity requires specialist counselling [18, 19]. In MSMDS, pregnancy in an affected woman carries high maternal risk and requires prior assessment [21].

9. Special Considerations: Population Diversity and Epidemiology

9.1. East Asia

Japan has the oldest data: in 2003, an estimated 7700 patients were under treatment (95% CI 6300–9300), with a female:male ratio of 1.8, peaks at 10–14 years in males and 20–24 years in females, 0.54 new diagnoses per 100,000 inhabitants per year (0.35 in 1994) and a family history in 12.1% [30]. The systematic review of population studies placed the Japanese incidence between 0.35 and 0.94 per 100,000 person-years [69], and the AHA/ASA cites a prevalence of 3.16/100,000 [18]. Orphanet records for Japan a prevalence of 6.2/100,000 and an incidence of 0.74/1,000,000 per year, both unvalidated [3]. In Korea, a national database recorded 8154 patients in 2011, a prevalence of 16.1/100,000, an incidence of 1.7–2.3/100,000 per year (2007–2011) and a female:male ratio of 1.8 [31]. In China, the prevalence in Nanjing was 3.92/100,000 (validated in Orphanet), the national incidence rose from 0.88/100,000 in 2016 to 1.47/100,000 in 2018, the highest prevalence was recorded in Henan (1.050/100,000), and the female:male ratio in a single-centre series was 1.5:1; in Taiwan, the incidence was much lower (0.02/100,000 person-years in the systematic review; 0.048/1,000,000 per year in Orphanet) [3, 34, 69]. The p.Arg4810Lys variant explains the higher Asian frequency: it is present in 0.5–2.5% of the general population of Japan, Korea and China and in about 80% of Japanese and Korean patients [20, 55, 58].

9.2. United States

In the states of Washington and California (1987–1998), the incidence was 0.086/100,000, with incidence ratios relative to white people of 4.6 (95% CI 3.4–6.3) in Asian Americans, 2.2 in African Americans and 0.5 (95% CI 0.3–0.8) in Hispanics [32]. In the national inpatient sample (2002–2008), the incidence was 0.57/100,000 per year, with a mean age of 32 years at diagnosis, a female:male ratio of 2.6 and 49% white, 24% black and 11% Asian patients [18]. In a recent four-state study (Florida, Georgia, Maryland and New York; 2011–2020; 4192 incident cases), the standardised incidence of MMA was 0.75/100,000 per year (MMD 0.57; MMS 0.18), higher in women (adjusted ratio 1.97), in Asians (2.46) and in black people (2.58) than in non-Hispanic white people, with an annual increase of 7.7%, particularly after 45 years of age [33]. These values measure incidence of hospital admission by diagnostic codes, whose validity for separating MMD from MMS is limited [18]. Up to 56% of Asian Americans with moyamoya have an RNF213 variant, versus 3.6–29% of non-Asian patients [36].

9.3. Europe

The ESO estimates that MMA is about 10 times less frequent in Western countries (0.047–0.086/100,000) than in East Asia, and probably underdiagnosed [2]. Orphanet records for France a point prevalence of 0.33/1,000,000 (expert opinion) and for Europe an incidence of 0.035/1,000,000 per year, unvalidated [3]. The French PNDS considers the incidence and prevalence in France imprecise and cites, in children, an incidence of 0.4/100,000 per year and a prevalence of 0.065/100,000, obtained by a survey of paediatric neurologists in 2011 [41]. European disease differs from Asian disease: less haemorrhage in adults (9.5% in a cohort of 200 patients), absence of p.Arg4810Lys and association with rare RNF213 variants in the RING domain, with GUCY1A1 (up to 2% of European patients) and with DIAPH1 [2, 11, 50, 58]. The sources consulted provide no epidemiological data from Portugal or Brazil; from Portugal there is the report, cited in GeneReviews, of a moyamoya pattern in a girl with Alagille syndrome and a JAG1 variant followed in Porto [26, 78].

9.4. Founder and consanguineous populations

p.Arg4810Lys is an East Asian founder allele with a shared haplotype [49]. The recessive forms (GUCY1A1, NOS3, the recessive ANO1 family) have been described in consanguineous families from France, Algeria, Morocco, Tunisia and Turkey, and one Mennonite family from Ontario had MMD in monozygotic twins and in a sibling [6, 11–13, 35].

9.5. Special populations

Children with trisomy 21, NF1, sickle cell disease, Alagille syndrome, MOPD II or cranial irradiation constitute populations at increased risk, in whom vascular imaging should be considered in the presence of symptoms and, in some, routinely [23, 36, 51]. MMA is associated with autoimmune diseases in white patients in the USA and with hyperthyroidism [1, 36].

9.6. European network and patient associations

The relevant European Reference Network is VASCERN (rare multisystemic vascular diseases), whose neurovascular disease working group (NEUROVASC-WG) includes moyamoya; the ESO guideline was endorsed by VASCERN [2, 44]. In France, the reference centre for rare vascular diseases of the brain and eye (CERVCO), the national reference centre for childhood stroke and the Association Tanguy Moyamoya are listed in the PNDS [41]. Orphanet maintains the entries for the entities of the class [3, 15–17].

10. Illustrative Clinical Case

10.1. Moyamoya with achalasia due to loss of function of GUCY1A1 (family F1 of Hervé and colleagues, 2014)

Two siblings, children of healthy consanguineous parents, with no relevant family or prenatal history, had repeated regurgitation in the first days of life; the oesophagogastroduodenal transit study showed megaoesophagus and manometry an achalasia pattern, and both underwent Heller myotomy, at 4 months and at 3 years [35]. The elder (F1 II-1) had ischaemic strokes at 3 and 9 years of age. Digital subtraction angiography showed occlusion of the terminal bifurcations of both ICAs and of the P2 segment of the right PCA, with a fine network of moyamoya vessels near the occluded arteries, that is, involvement of the anterior and posterior circulations (figure 2 of the article: https://pmc.ncbi.nlm.nih.gov/articles/PMC3951937/figure/fig2/) [35]. After the second stroke he was left with left hemiparesis and an opercular syndrome; the swallowing difficulties, from achalasia and from stroke, led to gastrostomy at 10 years and pneumatic dilatation of the cardia at 22; from 20 years of age he had hypertension controlled with nicardipine and perindopril; echocardiography and ultrasound of the kidneys and renal arteries were normal; platelets remained around 100,000/µL, with a bleeding time of 4 minutes and no clinical haemorrhages [35].

The brother (F1 II-3) had a different course: normal social and working life, mild intermittent dysphagia after the myotomy, extensive Raynaud phenomenon since childhood (fingers, hands, feet and nose), erectile dysfunction since adolescence and, at 25 years, malignant hypertension with severe retinopathy, posterior ischaemic optic neuropathy and renal failure (clearance of 50 mL/min), with small kidneys and normal renal arteries on ultrasound; at 28 years, the brain and intracranial arteries were normal [35]. Sanger sequencing of GUCY1A3 (now GUCY1A1) showed in both brothers the homozygous single-nucleotide deletion NM_000856.4:c.1170del, predicting p.(Glu391LysfsTer19) and designated with the same numbering on the MANE Select transcript NM_001130682.3 [35, 63]. The platelets of both lacked the α1 and β1 subunits of soluble guanylyl cyclase, aggregated less with ADP, adhered less to collagen and von Willebrand factor (55–63% of control) and did not respond to NO donors [35]. The pedigrees of the three families and the variants are shown in figure 1 of the article (https://pmc.ncbi.nlm.nih.gov/articles/PMC3951937/figure/fig1/) and the manometric recordings of the achalasia in figure 3 (https://pmc.ncbi.nlm.nih.gov/articles/PMC3951937/figure/fig3/) [35].

10.2. Lessons

The case shows the variable expressivity within the same family, with the same biallelic genotype: one brother with severe bilateral moyamoya, anterior and posterior, from the age of 3, and another without intracranial angiopathy at 28 years but with malignant hypertension and peripheral vascular phenomena [35]. Early achalasia, constant in this series, is the clue that should prompt GUCY1A1 testing and MR angiography before any digestive surgery, because moyamoya increases the perioperative ischaemic risk; in the sister from another family in the same series, sevoflurane anaesthesia was followed by an episode of microvascular dysfunction with cyanosis [35]. The platelet dysfunction did not translate into haemorrhages, which is relevant when antiplatelet therapy is being considered [35]. The treatment options proposed by the authors for the vascular symptoms (phosphodiesterase 5 inhibitors, NO donors) have not been tested and are not standard practice; cerebrovascular management follows the general recommendations for MMA [2, 35].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

Prognosis depends on age at onset, presentation, haemodynamic status, PCA involvement and access to revascularisation. In a symptomatic child without surgery, the risk of stroke has been estimated at 32% at 1 year and 66–90% at 5 years, falling to less than 5% after revascularisation in most series [36]. In Korea, 5-year survival was 92.9% in adults and 99.3% in children [31]. Intracranial haemorrhage is the main adverse prognostic factor and the main cause of death in MMA [40, 52]. PCA involvement and onset before 2–5 years of age are associated with more strokes, worse cognitive development and worse long-term social outcomes [2]. After combined surgery, late events are rare, but progression to the contralateral or posterior circulation continues (1.5% per patient-year) [2].

In the monogenic forms, prognosis is variable. RNF213 p.Arg4810Lys homozygotes and carriers of de novo RNF213 variants have early and severe disease [19, 20]. In MSMDS, 27% of the 33 patients died between 6 months and 32 years of age, and most known patients do not survive beyond the third decade, from aortic, cerebrovascular or pulmonary complications [21, 39]. In BRCC3 deletion, dilated cardiomyopathy may be the cause of death [9]. In MYMY6, severity ranges from severe disability in childhood to absence of angiopathy in young adulthood [35].

11.2. Surveillance

11.3. Rehabilitation and support

Motor and cognitive rehabilitation and psychological support are part of care at all stages, because of the risk and the disability associated with the disease [41]. In MSMDS, routine cognitive assessment guides rehabilitation and educational and home adaptations [21]. The treating physician detects neurological events and complications (stroke, TIA, headaches, cognitive deficits, epilepsy), prevents dehydration and the use of risky drugs and coordinates home care in the presence of disability [41].

12. Summary and Final Considerations

Moyamoya angiopathy is a progressive steno-occlusive disease of the terminal portion of the internal carotid arteries and their branches, with a network of fragile collaterals, which causes TIAs and infarcts mainly in children and haemorrhages and infarcts in adults. It is more frequent in East Asia, where the founder variant RNF213 p.(Arg4810Lys) is present in about 80% of patients and in 0.5–2.5% of the population, with low penetrance in the heterozygous state and high penetrance in the homozygous state. Outside Asia, the disease is about ten times less frequent, is associated with rare RNF213 variants in the RING domain and with a growing set of Mendelian genes: ACTA2 (with Arg179-related multisystemic smooth muscle dysfunction syndrome), GUCY1A1 and NOS3 of the nitric oxide pathway, BRCC3, ANO1, DIAPH1 and chromatin remodelling genes. The distinction between moyamoya disease and syndrome is nosological; the most frequent systemic genetic diseases with MMS are NF1, trisomy 21 and sickle cell disease.

The diagnosis is radiological. The 2021 Japanese criteria accept unilateral disease, require angiography when there is doubt about atherosclerosis and add the reduction in outer arterial diameter on MRI. Genetic testing does not replace imaging, has selective indications (early onset, family history, consanguinity, systemic signs, extracranial disease, East Asian ancestry) and should be interpreted with care: p.Arg4810Lys is a risk allele and not a Mendelian pathogenic variant, and a negative result does not exclude a genetic cause. Simple clinical clues guide testing: congenital mydriasis with patent ductus arteriosus, early achalasia, hypogonadism with short stature in a male, posterior circulation aneurysms and extracranial stenoses in an infant.

Surgical revascularisation, direct, indirect or combined, is the central treatment in symptomatic disease; the only randomised trial, JAM, showed a reduction in rebleeding with direct bypass in Japanese adults. Antiplatelet therapy, with aspirin as first line, is suggested in non-haemorrhagic disease on very low quality evidence, and the European, Japanese, American, Chinese and French guidelines diverge on genetic testing, haemodynamic assessment, thrombolysis, and the technique and timing of surgery. Prevention of hyperventilation, dehydration and hypotension, planning of anaesthesia and avoidance of vasoconstrictors are measures common to all guidelines. There is no approved targeted therapy; nicotinamide riboside in MSMDS is in a phase 1 trial after preclinical results, and sapropterin has only an n-of-1 report. Imaging follow-up should be prolonged, because the disease progresses even after surgery, and family counselling depends on the cause identified, with selective screening of relatives and cascade testing in the Mendelian forms. The VASCERN network and the national referral centres concentrate the surgical experience on which outcomes depend.

13. References

1. Kuroda S, Fujimura M, Takahashi J, Kataoka H, Ogasawara K, Iwama T, et al. Diagnostic Criteria for Moyamoya Disease - 2021 Revised Version. Neurol Med Chir (Tokyo). 2022;62(7):307-312. https://doi.org/10.2176/jns-nmc.2022-0072
2. Bersano A, Khan N, Fuentes B, Acerbi F, Canavero I, Tournier-Lasserve E, et al. European Stroke Organisation (ESO) Guidelines on Moyamoya angiopathy Endorsed by Vascular European Reference Network (VASCERN). Eur Stroke J. 2023;8(1):55-84. https://doi.org/10.1177/23969873221144089
3. Orphanet. Moyamoya disease (ORPHA:2573) [Internet]. Paris: INSERM; accessed 31/10/2026. https://www.orpha.net/en/disease/detail/2573
4. Orphanet. Moyamoya angiopathy (ORPHA:477768) [Internet]. Paris: INSERM; accessed 31/10/2026. https://www.orpha.net/en/disease/detail/477768
5. Orphanet. Radiation-induced Moyamoya syndrome (ORPHA:726832) [Internet]. Paris: INSERM; accessed 31/10/2026. https://www.orpha.net/en/disease/detail/726832
6. Online Mendelian Inheritance in Man, OMIM®. %252350 Moyamoya disease 1; MYMY1 [Internet]. Baltimore (MD): Johns Hopkins University; accessed 31/10/2026. https://omim.org/entry/252350
7. OMIM®. #607151 Moyamoya disease 2, susceptibility to; MYMY2 [Internet]. Accessed 31/10/2026. https://omim.org/entry/607151
8. OMIM®. %608796 Moyamoya disease 3; MYMY3 [Internet]. Accessed 31/10/2026. https://omim.org/entry/608796
9. OMIM®. #300845 Moyamoya disease 4 with short stature, hypergonadotropic hypogonadism, and facial dysmorphism; MYMY4 [Internet]. Accessed 31/10/2026. https://omim.org/entry/300845
10. OMIM®. #614042 Moyamoya disease 5; MYMY5 [Internet]. Accessed 31/10/2026. https://omim.org/entry/614042
11. OMIM®. #615750 Moyamoya disease 6 with or without achalasia; MYMY6 [Internet]. Accessed 31/10/2026. https://omim.org/entry/615750
12. OMIM®. #620687 Moyamoya disease 7; MYMY7 [Internet]. Accessed 31/10/2026. https://omim.org/entry/620687
13. OMIM®. #621469 Moyamoya disease 8; MYMY8 [Internet]. Accessed 31/10/2026. https://omim.org/entry/621469
14. OMIM®. #613834 Multisystemic smooth muscle dysfunction syndrome; MSMDS [Internet]. Accessed 31/10/2026. https://omim.org/entry/613834
15. Orphanet. Moyamoya disease with early-onset achalasia (ORPHA:401945) [Internet]. Paris: INSERM; accessed 31/10/2026. https://www.orpha.net/en/disease/detail/401945
16. Orphanet. Moyamoya angiopathy-short stature-facial dysmorphism-hypergonadotropic hypogonadism syndrome (ORPHA:280679) [Internet]. Paris: INSERM; accessed 31/10/2026. https://www.orpha.net/en/disease/detail/280679
17. Orphanet. Multisystemic smooth muscle dysfunction syndrome (ORPHA:404463) [Internet]. Paris: INSERM; accessed 31/10/2026. https://www.orpha.net/en/disease/detail/404463
18. Gonzalez NR, Amin-Hanjani S, Bang OY, Coffey C, Du R, Fierstra J, et al. Adult Moyamoya Disease and Syndrome: Current Perspectives and Future Directions: A Scientific Statement From the American Heart Association/American Stroke Association. Stroke. 2023;54(10):e465-e479. https://doi.org/10.1161/STR.0000000000000443
19. 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):803-10. https://doi.org/10.1212/WNL.0b013e318249f71f
20. Pinard A, Fiander MDJ, Cecchi AC, Rideout AL, Azouz M, Fraser SM, et al. Association of De Novo RNF213 Variants With Childhood Onset Moyamoya Disease and Diffuse Occlusive Vasculopathy. Neurology. 2021;96(13):e1783-e1791. https://doi.org/10.1212/WNL.0000000000011653
21. Regalado ES, Mellor-Crummey L, De Backer J, Braverman AC, Ades L, Benedict S, et al. Clinical history and management recommendations of the smooth muscle dysfunction syndrome due to ACTA2 arginine 179 alterations. Genet Med. 2018;20(10):1206-1215. https://doi.org/10.1038/gim.2017.245
22. Milewicz DM, Cecchi AC. Heritable Thoracic Aortic Disease Overview. 2003 Feb 13 [updated 2023 May 4]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK1120/
23. Duker A, Jackson A, Bober MB. Microcephalic Osteodysplastic Primordial Dwarfism Type II. 2021 Dec 30 [revised 2025 Nov 20]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK575926/
24. Crow YJ. Aicardi-Goutières Syndrome. 2005 Jun 29 [updated 2016 Nov 22]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK1475/
25. Friedman JM. Neurofibromatosis 1. 1998 Oct 2 [updated 2022 Apr 21; revised 2025 Apr 3]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK1109/
26. Spinner NB, Loomes KM, Krantz ID, Gilbert MA. Alagille Syndrome. 2000 May 19 [updated 2024 Jan 4]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK1273/
27. Bender MA, Carlberg K. Sickle Cell Disease. 2003 Sep 15 [updated 2022 Nov 17; revised 2025 Feb 13]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK1377/
28. Lippner E, Lücke T, Salgado C, Boerkoel C, Lewis DB. Schimke Immunoosseous Dysplasia. 2002 Oct 1 [updated 2022 Apr 14]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK1376/
29. Mehta PA, Ebens CL. Fanconi Anemia. 2002 Feb 14 [updated 2026 Jan 15]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK1401/
30. Kuriyama S, Kusaka Y, Fujimura M, Wakai K, Tamakoshi A, Hashimoto S, et al. Prevalence and clinicoepidemiological features of moyamoya disease in Japan: findings from a nationwide epidemiological survey. Stroke. 2008;39(1):42-7. https://doi.org/10.1161/STROKEAHA.107.490714
31. Ahn IM, Park DH, Hann HJ, Kim KH, Kim HJ, Ahn HS. Incidence, prevalence, and survival of moyamoya disease in Korea: a nationwide, population-based study. Stroke. 2014;45(4):1090-5. https://doi.org/10.1161/STROKEAHA.113.004273
32. Uchino K, Johnston SC, Becker KJ, Tirschwell DL. Moyamoya disease in Washington State and California. Neurology. 2005;65(6):956-8. https://doi.org/10.1212/01.wnl.0000176066.33797.82
33. Miller R, Anikpezie N, Hoffman H, Bako AT, Patel SD, Sahoo A, et al. Demographic Disparities in the Incidence of Moyamoya Angiopathy in the United States. Neurology. 2025;105(5):e214013. https://doi.org/10.1212/WNL.0000000000214013
34. Ni W, Su J, Fei Y, Gu Y, Mao Y, Wang S. Chinese expert consensus on the treatment and diagnosis of moyamoya disease and moyamoya syndrome (2024). Stroke Vasc Neurol. 2026;11(4):406-414. https://doi.org/10.1136/svn-2024-003925
35. Hervé D, Philippi A, Belbouab R, Zerah M, Chabrier S, Collardeau-Frachon S, et al. Loss of α1β1 soluble guanylate cyclase, the major nitric oxide receptor, leads to moyamoya and achalasia. Am J Hum Genet. 2014;94(3):385-94. https://doi.org/10.1016/j.ajhg.2014.01.018
36. Ferriero DM, Fullerton HJ, Bernard TJ, Billinghurst L, Daniels SR, DeBaun MR, et al. Management of Stroke in Neonates and Children: A Scientific Statement From the American Heart Association/American Stroke Association. Stroke. 2019;50(3):e51-e96. https://doi.org/10.1161/STR.0000000000000183
37. Miskinyte S, Butler MG, Hervé D, Sarret C, Nicolino M, Petralia JD, et al. Loss of BRCC3 deubiquitinating enzyme leads to abnormal angiogenesis and is associated with syndromic moyamoya. Am J Hum Genet. 2011;88(6):718-728. https://doi.org/10.1016/j.ajhg.2011.04.017
38. Pinard A, Ye W, Fraser SM, Rosenfeld JA, Pichurin P, Hickey SE, et al. Rare variants in ANO1, encoding a calcium-activated chloride channel, predispose to moyamoya disease. Brain. 2023;146(9):3616-3623. https://doi.org/10.1093/brain/awad172
39. Mulder BJM, van de Laar IMBH, De Backer J. Heritable Thoracic Aortic Diseases: Syndromal and Isolated (F)TAAD. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 309-43. https://doi.org/10.1007/978-3-030-45457-9_19
40. Fujimura M, Tominaga T, Kuroda S, Takahashi JC, Endo H, Ogasawara K, et al. 2021 Japanese Guidelines for the Management of Moyamoya Disease: Guidelines from the Research Committee on Moyamoya Disease and Japan Stroke Society. Neurol Med Chir (Tokyo). 2022;62(4):165-170. https://doi.org/10.2176/jns-nmc.2021-0382
41. Haute Autorité de Santé; CERVCO. Protocole National de Diagnostic et de Soins (PNDS): Maladie et syndrome de moyamoya de l'enfant et de l'adulte — synthèse à destination du médecin traitant [Internet]. Saint-Denis: HAS; 2017. https://has-sante.fr/upload/docs/application/pdf/2017-03/dir3/pnds_76_-_maladie_et_syndrome_de_moyamoya_de_lenfant_et_de_ladulte_-_synthese_a_destination_du_medecin_traitant.pdf
42. Li Q, Cui L, Su J, Shen Y. Case report: Multisystemic smooth muscle dysfunction syndrome: a rare genetic cause of infantile interstitial lung disease. Front Pharmacol. 2025;15:1510969. https://doi.org/10.3389/fphar.2024.1510969
43. Kaw A, Kaw K, Hostetler EM, Beleza-Meireles A, Smith-Collins A, Armstrong C, et al. Expanding ACTA2 genotypes with corresponding phenotypes overlapping with smooth muscle dysfunction syndrome. Am J Med Genet A. 2022;188(8):2389-2396. https://doi.org/10.1002/ajmg.a.62775
44. VASCERN — European Reference Network on Rare Multisystemic Vascular Diseases. Neurovascular Diseases Working Group (NEUROVASC-WG) [Internet]. Accessed 31/10/2026. https://vascern.eu/group/neurovascular-diseases-neurovasc-wg/
45. Pinard A, Guey S, Guo D, Cecchi AC, Kharas N, Wallace S, et al. The pleiotropy associated with de novo variants in CHD4, CNOT3, and SETD5 extends to moyamoya angiopathy. Genet Med. 2020;22(2):427-431. https://doi.org/10.1038/s41436-019-0639-2
46. Guo DC, Papke CL, Tran-Fadulu V, Regalado ES, Avidan N, Johnson RJ, et al. Mutations in smooth muscle alpha-actin (ACTA2) cause coronary artery disease, stroke, and Moyamoya disease, along with thoracic aortic disease. Am J Hum Genet. 2009;84(5):617-27. https://doi.org/10.1016/j.ajhg.2009.04.007
47. Milewicz DM, Østergaard JR, Ala-Kokko LM, Khan N, Grange DK, Mendoza-Londono R, et al. De novo ACTA2 mutation causes a novel syndrome of multisystemic smooth muscle dysfunction. Am J Med Genet A. 2010;152A(10):2437-43. https://doi.org/10.1002/ajmg.a.33657
48. 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. https://doi.org/10.1038/jhg.2010.132
49. 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. https://doi.org/10.1371/journal.pone.0022542
50. Kundishora AJ, Peters ST, Pinard A, Duran D, Panchagnula S, Barak T, et al. DIAPH1 Variants in Non-East Asian Patients With Sporadic Moyamoya Disease. JAMA Neurol. 2021;78(8):993-1003. https://doi.org/10.1001/jamaneurol.2021.1681
51. Ullrich NJ, Robertson R, Kinnamon DD, Scott RM, Kieran MW, Turner CD, et al. Moyamoya following cranial irradiation for primary brain tumors in children. Neurology. 2007;68(12):932-8. https://doi.org/10.1212/01.wnl.0000257095.33125.48
52. Rifino N, Hervè D, Acerbi F, Kuroda S, Lanzino G, Vajkoczy P, et al. Diagnosis and management of adult Moyamoya angiopathy: An overview of guideline recommendations and identification of future research directions. Int J Stroke. 2025;20(5):512-523. https://doi.org/10.1177/17474930241297031
53. Lauer A, Speroni SL, Patel JB, Regalado E, Choi M, Smith E, et al. Cerebrovascular Disease Progression in Patients With ACTA2 Arg179 Pathogenic Variants. Neurology. 2021;96(4):e538-e552. https://doi.org/10.1212/WNL.0000000000011210
54. Santoro C, Giugliano T, Kraemer M, Torella A, Schwitalla JC, Cirillo M, et al. Whole exome sequencing identifies MRVI1 as a susceptibility gene for moyamoya syndrome in neurofibromatosis type 1. PLoS One. 2018;13(7):e0200446. https://doi.org/10.1371/journal.pone.0200446
55. GeneGuidance. RNF213 — clinical curation of the gene and the condition Moyamoya disease 2, susceptibility to (OMIM 607151) [local database]. Accessed 31/10/2026.
56. VariantValidator. NM_001256071.3:c.14429G>A (RNF213) — validation and GRCh38 genomic mapping [Internet]. Manchester: University of Manchester; accessed 31/10/2026. https://variantvalidator.org/
57. National Center for Biotechnology Information. ClinVar: VCV000039700, NM_001256071.3(RNF213):c.14429G>A (p.Arg4810Lys) [Internet]. Bethesda (MD): NCBI; accessed 31/10/2026. https://www.ncbi.nlm.nih.gov/clinvar/variation/39700/
58. Guey S, Kraemer M, Hervé D, Ludwig T, Kossorotoff M, Bergametti F, et al. Rare RNF213 variants in the C-terminal region encompassing the RING-finger domain are associated with moyamoya angiopathy in Caucasians. Eur J Hum Genet. 2017;25(8):995-1003. https://doi.org/10.1038/ejhg.2017.92
59. GeneGuidance. ACTA2 — clinical curation of the gene [local database]. Accessed 31/10/2026.
60. VariantValidator. NM_001613.4:c.536G>A (ACTA2) [Internet]. Accessed 31/10/2026. https://variantvalidator.org/
61. Kaw A, Majumder S, Esparza Pinelo JE, Wu T, Starosolski Z, Zhou Z, et al. Immature Acta2(R179C/+) smooth muscle cells cause moyamoya-like cerebrovascular lesions in mice prevented by boosting OXPHOS. Nat Commun. 2025;16(1):6105. https://doi.org/10.1038/s41467-025-61042-3
62. Erdmann J, Schunkert H. Genetics of (Premature) Coronary Artery Disease. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 413-30. https://doi.org/10.1007/978-3-030-45457-9_24
63. VariantValidator. GUCY1A1 NM_000856.4 e NM_001130682.3 (MANE): c.1045C>T, c.1170del, c.1086+1G>A [Internet]. Accessed 31/10/2026. https://variantvalidator.org/
64. OMIM®. *604673 Inositol 1,4,5-triphosphate receptor associated 1; IRAG1 [Internet]. Accessed 31/10/2026. https://omim.org/entry/604673
65. GeneGuidance. BRCC3 — Moyamoya angiopathy-short stature-facial dysmorphism-hypergonadotropic hypogonadism syndrome [local database]. Accessed 31/10/2026.
66. GeneGuidance. DIAPH1 — clinical curation of the gene [local database]. Accessed 31/10/2026.
67. Weiss K, Lachlan K. CHD4 Neurodevelopmental Disorder. 2020 Sep 3. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK561516/
68. Ziegler SG, Ferreira CR. Generalized Arterial Calcification of Infancy. 2014 Nov 13 [updated 2026 May 19]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK253403/
69. Kleinloog R, Regli L, Rinkel GJ, Klijn CJ. Regional differences in incidence and patient characteristics of moyamoya disease: a systematic review. J Neurol Neurosurg Psychiatry. 2012;83(5):531-6. https://doi.org/10.1136/jnnp-2011-301387
70. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 884-85.
71. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 796-99.
72. Kainth DS, Chaudhry SA, Kainth HS, Suri FK, Qureshi AI. Prevalence and characteristics of concurrent down syndrome in patients with moyamoya disease. Neurosurgery. 2013;72(2):210-5; discussion 215. https://doi.org/10.1227/NEU.0b013e31827b9beb
73. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 716-19.
74. Dobson SR, Holden KR, Nietert PJ, Cure JK, Laver JH, Disco D, et al. Moyamoya syndrome in childhood sickle cell disease: a predictive factor for recurrent cerebrovascular events. Blood. 2002;99(9):3144-50. https://doi.org/10.1182/blood.v99.9.3144
75. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 228-31.
76. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 128-31.
77. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 822-25.
78. Rocha R, Soro I, Leitão A, Silva ML, Leão M. Moyamoya vascular pattern in Alagille syndrome. Pediatr Neurol. 2012;47(2):125-128. https://doi.org/10.1016/j.pediatrneurol.2012.04.014
79. VariantValidator. NM_000214.3:c.715dup (JAG1) — validation and predicted protein consequence [Internet]. Manchester: University of Manchester; accessed 31/10/2026. https://variantvalidator.org/
80. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 39, Disorders of Nucleic Acid Metabolism, tRNA Metabolism and Ribosomal Biogenesis, section 39.1.2.
81. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 218-21.
82. Vega H, Gordillo M, Jabs EW. ESCO2 Spectrum Disorder. 2006 Apr 18 [updated 2025 Aug 14]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK1153/
83. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 424-27.
84. Izumi K, Raible S, Krantz I. AFF4-Related CHOPS Syndrome. 2026 Feb 19. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK621037/
85. Smith ACM, Berens J, Boyd KE, Brennan C, Gropman A, Haas-Givler B, et al. Smith-Magenis Syndrome. 2001 Oct 22 [updated 2025 May 29]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK1310/
86. Stevenson D, Legius E. Legius Syndrome. 2010 Oct 14 [updated 2025 Dec 4]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK47312/
87. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 616-17.
88. VariantValidator. Validation of NM_001613.4 (ACTA2): c.115C>T, c.137T>G, c.197T>A, c.535C>A, c.535C>G, c.535C>T, c.536G>T, c.592C>T e c.611C>T [Internet]. Manchester: University of Manchester; accessed 01/11/2026. https://variantvalidator.org/
89. Rahman AA, Stavely R, Ott LC, Han CY, Ohishi K, Hotta R, et al. Structural and functional gastrointestinal abnormalities in ACTA2 R179H mice modeling multisystemic smooth muscle dysfunction syndrome. JCI Insight. 2026;11(4):e190469. https://doi.org/10.1172/jci.insight.190469
90. Lupo V, Di Gregorio MG, Mastrogiorgio G, Magliozzi M, Scapillati ME, Maglione V, et al. Neonatal diagnosis of ACTA2-related disease: A case report and review of literature. Am J Med Genet A. 2023;191(4):1111-1118. https://doi.org/10.1002/ajmg.a.63118
91. Roylance A, Spinty S, Pettorini B. Communicating hydrocephalus and raised intracranial pressure in association with multi-systemic smooth muscle dysfunction syndrome (MSMDS). Childs Nerv Syst. 2023;39(3):825-828. https://doi.org/10.1007/s00381-022-05703-4
92. Cuoco JA, Busch CM, Klein BJ, Benko MJ, Stein R, Nicholson AD, et al. ACTA2 Cerebral Arteriopathy: Not Just a Puff of Smoke. Cerebrovasc Dis. 2018;46(3-4):161-171. https://doi.org/10.1159/000493863
93. Madjidyar J, Kyselyova A, Kulcsar Z, Schubert T, Thurner P, Khan N. Moyamoya-Type Collaterals in ACTA2 p. Arg179 Related Arteriopathy: Challenging the Paradigm of Absent Neovascularization-A Longitudinal DSA Study. AJNR Am J Neuroradiol. 2026 [publicação eletrónica antes da impressão: 12/10/2026]. https://doi.org/10.3174/ajnr.A9674
94. Focke JK, Kraemer M. A familial missense ACTA2 variant p.Arg198Cys leading to Moyamoya-like arteriopathy with straight course of the intracranial arteries, aortic aneurysm and lethal aortic dissection. Neurol Res Pract. 2023;5(1):38. https://doi.org/10.1186/s42466-023-00268-2
95. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, Clinical Approach to Inborn Errors of Metabolism in Paediatrics, section 1.4.1 (Table 1.6).
96. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 137-41.
97. El-Hattab AW, Almannai M, Scaglia F. MELAS. 2001 Feb 27 [updated 2018 Nov 29]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. https://www.ncbi.nlm.nih.gov/books/NBK1233/
98. Koss M, Scott RM, Irons MB, Smith ER, Ullrich NJ. Moyamoya syndrome associated with neurofibromatosis Type 1: perioperative and long-term outcome after surgical revascularization. J Neurosurg Pediatr. 2013;11(4):417-25. https://doi.org/10.3171/2012.12.PEDS12281
99. GeneGuidance. SAMHD1 — clinical curation of the gene [local database]. Accessed 31/10/2026.
100. Marie S, Dewulf JP, Nassogne MC. Disorders of Purine and Pyrimidine Metabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. p. 587-614 (section 32.3.2). https://doi.org/10.1007/978-3-662-63123-2_32
101. Monagle P, Chan AKC, Goldenberg NA, Ichord RN, Journeycake JM, Nowak-Göttl U, et al. Antithrombotic therapy in neonates and children: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. 2012;141(2 Suppl):e737S-e801S. https://doi.org/10.1378/chest.11-2308
102. INFARMED. Summary of Product Characteristics: Aspirina GR 100 mg gastro-resistant tablets (Bayer Portugal) [Internet]. Lisbon: INFARMED; revision 04/2026. https://extranet.infarmed.pt/INFOMED-fo/
103. ClinicalTrials.gov. Study on the Effects of Pharmacological and Physical Therapy in Delaying the Progression of Moyamoya Disease (Moyamoya Syndrome). NCT07762547 [Internet]. Updated 13/08/2026; accessed 31/10/2026. https://clinicaltrials.gov/study/NCT07762547
104. INFARMED. Summary of Product Characteristics: Cilostazol Ferrer 100 mg tablets [Internet]. Lisbon: INFARMED; renewal 05/06/2023. https://extranet.infarmed.pt/INFOMED-fo/
105. Morita S, Yamaguchi K, Akagawa H, Ishikawa T, Funatsu T, Eguchi S, et al. Triple bypass for multisystem smooth muscle dysfunction syndrome due to Arg179His ACTA2 mutation. J Stroke Cerebrovasc Dis. 2022;31(9):106402. https://doi.org/10.1016/j.jstrokecerebrovasdis.2022.106402
106. Muroi A, Shiono J, Ihara S, Morisaki H, Nakai Y. Nonsurgical treatment of cerebral ischemia associated with ACTA2 cerebral arteriopathy: a case report and literature review. Childs Nerv Syst. 2022;38(6):1209-1212. https://doi.org/10.1007/s00381-021-05360-z
107. European Medicines Agency. Tracleer (bosentan): Summary of Product Characteristics [Internet]. Amsterdam: EMA; initial authorisation 15/05/2002, last renewal 20/04/2012; accessed 01/11/2026. https://www.ema.europa.eu/en/documents/product-information/tracleer-epar-product-information_en.pdf
108. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 59-61.
109. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 506-11.
110. Takahashi JC, Ikeda T, Iihara K, Miyamoto S. Pregnancy and delivery in moyamoya disease: results of a nationwide survey in Japan. Neurol Med Chir (Tokyo). 2012;52(5):304-10. https://doi.org/10.2176/nmc.52.304
111. Maragkos GA, Ascanio LC, Chida K, Boone MD, Ogilvy CS, Thomas AJ, et al. Moyamoya disease in pregnancy: a systematic review. Acta Neurochir (Wien). 2018;160(9):1711-1719. https://doi.org/10.1007/s00701-018-3597-6
112. Hausman-Kedem M, Bielopolski N, Krishnan V, Wald-Altman S, Shrian SI, Bar-Yosef O, et al. Re-Purposing Sapropterin (Kuvan) for ACTA2-Related Multisystemic Smooth Muscle Dysfunction Syndrome: A Translational Mechanistic and First-In-Human Therapeutic Report. Ann Clin Transl Neurol. 2026 [publicação eletrónica antes da impressão: 26/07/2026]. https://doi.org/10.1002/acn3.70493
113. ClinicalTrials.gov. Nicotinamide Riboside (NR) to Treat Moyamoya-like Cerebrovascular Disease in Smooth Muscle Dysfunction Syndrome (SMDS). NCT06280482 [Internet]. Bethesda (MD): NLM; updated 12/09/2025; accessed 31/10/2026. https://clinicaltrials.gov/study/NCT06280482
114. Huang SF, Xu JL, Ren CH, Sim N, Han C, Han YQ, et al. Remote ischemic conditioning prevents ischemic cerebrovascular events in children with moyamoya disease: a randomized controlled trial. World J Pediatr. 2024;20(9):925-934. https://doi.org/10.1007/s12519-024-00824-z
115. ClinicalTrials.gov. Search «moyamoya» (67 studies) [Internet]. Accessed 31/10/2026. https://clinicaltrials.gov/search?cond=moyamoya
116. Ding Q, Gan P, Xu Z, Li H, Guo L, MacDonald C, et al. Genomic Editing of a Pathogenic Sequence Variant in ACTA2 Rescues Multisystemic Smooth Muscle Dysfunction Syndrome in Mice. Circulation. 2025;152(7):465-483. https://doi.org/10.1161/CIRCULATIONAHA.125.074218
117. Alves CRR, Das S, Krishnan V, Ha LL, Fox LR, Stutzman HE, et al. Treatment of a severe vascular disease using a bespoke CRISPR-Cas9 base editor in mice. Nat Biomed Eng. 2026;10(5):952-967. https://doi.org/10.1038/s41551-025-01499-1
118. ClinicalTrials.gov. A Retrospective and Prospective Natural History of Genetic Vasculopathies. NCT06552052 [Internet]. Bethesda (MD): NLM; accessed 01/11/2026. https://clinicaltrials.gov/study/NCT06552052
119. ClinicalTrials.gov. A Single Patient Study of Sapropterin for Multisystem Smooth Muscle Dysfunction Syndrome. NCT07574034 [Internet]. Bethesda (MD): NLM; accessed 01/11/2026. https://clinicaltrials.gov/study/NCT07574034