Introduction

Moyamoya disease is an idiopathic, bilateral progressive stenotic vasculopathy with the resultant formation of fragile, collateral “moyamoya” vessels [22, 24]. Unilateral disease or disease occurring secondary to known predisposing conditions such as neurofibromatosis, Down’s syndrome, radiation, and sickle cell disease, among others, are referred to as moyamoya syndrome [22]. Both moyamoya disease and moyamoya syndrome can serve as a considerable source of neurological morbidity as a result of recurrent ischemic events, and, in adults, hemorrhage [7, 10, 22, 24]. Ischemic events in both are known to be mitigated by direct revascularization [8–12] and, in children, indirect revascularization as well [6, 22, 23]. These neurosurgical tenets have withstood the test of time, in light of the results of the original extracranial-intracranial (EC-IC) bypass study [2] and the more recent Carotid Occlusion Surgery Study (COSS) study [19]. Although results of SAMMPRIS may mitigate endovascular approaches to cerebral ischemic disease [1], results of this study do not apply to moyamoya as it is pathogenetically different from atheromatous disease. Rather, this specific phenomenon should be evaluated separately. Currently, there is only a paucity of reports detailing endovascular approaches to moyamoya [3, 4, 13, 14, 18, 20, 21]. It is important to address whether this is merely a result of a bias for proven surgical revascularization methods or whether it is also a result of a lack of efficacy of these approaches. In this manuscript, we comprehensively review the literature on endovascular approaches to moyamoya, incorporating a case of our own while we evaluate treatment approach (angioplasty and/or stent), complications, and both angiographic and clinical outcomes at last follow-up.

Methods

We performed a comprehensive review of the literature querying the PubMed and Embase databases with the search terms, “moyamoya,” “endovascular,” “stent,” “wingspan,” and “angioplasty.” All reports describing angioplasty and/or stenting of stenotic intracranial vessels in a patient with moyamoya were collected. One report that described stenting of an atherosclerotic, stenosed external carotid artery in a patient with moyamoya was thus excluded [15].

We extracted patient age stratified into pediatric (age less than 18) or adult, pathology type (moyamoya disease vs moyamoya syndrome) treatment modality (angioplasty alone vs stent), vessel treated, procedural complications, and angiographic and clinical results at last follow-up. We also incorporated an attempt of our own among these results, representing 1 attempted endovascular approach in our last 50 consecutive patients seen by our neurosurgical service with moyamoya. Results were initially pooled and then segregated by approach (stenting vs angioplasty alone). All comparisons were performed employing the Fisher exact test. Monthly rates of angiographic and/or clinical recurrence were calculated by dividing the number of angiographic recurrences by the follow-up period (in months). This follow-up period concluded at the time of surgical revascularization, if performed. A clinical recurrence was defined by at least one recurrent transient ischemic attack or stroke after treatment.

Results

We collected results across a total of seven reports [3, 4, 13, 14, 18, 20, 21], adding our own case. Two of these were North American case series, the remainder single case reports. Notably, one of the case series included patients with moyamoya disease [13] while the other incorporated patients with moyamoya syndrome [18]. Overall, a total of 17 patients underwent 28 endovascular procedures (11 stenting, 17 angioplasty alone). Six patients had moyamoya disease and 11 had moyamoya syndrome (3 with idiopathic unilateral disease, 6 secondary to atherosclerosis, 1 child secondary to essential thrombocythemia, and another child secondary to a complex congenital syndromic condition). The vessel treated was the internal carotid artery (ICA) in 11 cases and the middle cerebral artery, first segment (M1) in 17 cases. All patients had symptomatic ischemic disease, and all but two were adults.

Overall, in three procedures, the vessel could not be effectively dilated (11 %), and after two, the procedure was complicated by clinically devastating intracerebral hemorrhage (7 %). Two procedural failures occurred in patients with moyamoya syndrome (11 % of procedures) and one in a patient with moyamoya disease (11 % of procedures, p = 1.0 as compared to moyamoya syndrome). One hemorrhage occurred in a patient with moyamoya syndrome (5 % of procedures) and another in a patient with moyamoya disease (11 % of procedures, p = 1.0 as compared to moyamoya syndrome).

For the remaining 23 procedures, the mean subsequent follow-up was 7.0 months, corresponding to a total of 161.5 patient-months. Angiographic recurrence was seen in 16 cases (70 %) while recurrence of clinical symptoms was seen in 13 cases (57 %). The monthly rates of each of these were 9.3 and 8.0 %, respectively. Angiographic recurrence occurred after 5/7 procedures performed in patients with moyamoya disease (71 %) and in 11/16 performed in patients with moyamoya syndrome (69 %, p = 1.0). Clinical recurrence occurred after 5/7 procedures performed in patients with moyamoya disease (71 %) and in 8/16 performed in patients with moyamoya syndrome (50 %, p = 0.41). Procedural success, defined as a lack of both angiographic and clinical recurrence at follow-up, was achieved after seven procedures (25 %); five performed in patients with moyamoya syndrome (26 % of procedures) and two in patients with moyamoya disease (22 % of procedures, p = 1.0). Of 15 patients that did not suffer devastating intracranial hemorrhages from their procedure, 8 ultimately underwent surgical revascularization (53 %).

Results stratified by treatment modality are provided in Tables 1 and 2. Twelve patients underwent 17 angioplasty procedures (Table 1) while 9 patients underwent 11 stenting procedures (Table 2). Of note, 10 of 11 stents deployed were Wingspan [3, 13, 18, 21], while the other was an AVE INX3 coronary stent [14]. There was no statistically significant difference between the angioplasty alone and stenting groups in the rate of post-procedural hemorrhage (6 % for angioplasty vs 9 % for stenting, p = 1.0), angiographic recurrence (66 % for angioplasty vs 70 % for stenting, p = 1.0), or clinical recurrence (54 % for angioplasty vs 60 % for stenting, p = 1.0). Technical procedural failure was not seen in any of the stenting cases, while it was seen in 3/17 angioplasty procedures (18 %, p = 0.26 vs stenting). Procedural success, defined as a lack of both angiographic and clinical recurrence at follow-up, was achieved after 24 % of angioplasty procedures and after 27 % of stenting cases (p = 1.0).

Table 1 Cases undergoing angioplasty alonea
Full size table
Table 2 Cases undergoing stent placementa
Full size table

Discussion

Early enthusiasm for the endovascular management of non-acute cerebrovascular ischemic disease [16, 26] was tempered by the recent results of the SAMMPRIS trial [1], demonstrating greater rates of stroke or death at 30 days and 1 year after Wingspan stenting as compared to medical management of severe intracranial stenosis. However, these results do not apply to moyamoya given its significant pathophysiologic difference from atheromatous disease [5, 22]. More germane to moyamoya patients, a separate study evaluating patterns of Wingspan stent restenosis found significantly greater rates in younger patients and those with supraclinoid ICA stenosis [25].

Although a lack of efficacy was seen for surgical revascularization in patients with cerebral ischemic disease in the original EC-IC bypass [2] and COSS studies [1], results of these studies strongly contrast with the litany of reports describing the efficacy of surgical revascularization for moyamoya [6, 8–12, 23]. This review illustrates the remarkable lack of such studies evaluating endovascular approaches, a topic that is still reportable in a case report format [3, 4, 21]. Although the known success of surgical revascularization may limit consideration of endovascular approaches, the relative potential technical ease of endovascular therapies may have catalyzed an espousal of these approaches if they were efficacious. We feel the lack of reports detailing endovascular management of moyamoya underscores the limited efficacy of these approaches, as unsuccessful attempts such as our contributed case are less likely to be reported. Indeed, one of our incorporated case series comes from a tertiary referral center for surgical revascularization [13]. This center reported prior outside hospital endovascular results that preceded its own surgical revascularization. Interestingly, this series illustrated the efficacy of surgical revascularization after failed attempted endovascular treatment, as all patients were asymptomatic after direct revascularization procedures after a mean follow-up of 18 months [13].

It is noteworthy that the procedural success rate in our study, defined as freedom from angiographic and clinical recurrence, was only 25 %, almost a similar number to the procedural complication and failure rate (18–7 % hemorrhage rate, 11 % failure to dilate vessel rate). Intuitively, the success rate is also influenced by the follow-up period, which was frequently very limited, thus potentially inflating the actual procedural success rate. Although procedural risk other than hemorrhage was not clearly provided in all studies, it is likely in itself quite significant. Remarkably, 53 % of patients that did not suffer from hemorrhagic complications of the procedure ultimately underwent surgical revascularization anyhow. Most importantly, the rates of angiographic and clinical recurrence were exceedingly high, requiring analysis and reporting at a monthly rate—9.3 and 8.0 %, respectively. The relative lack of efficacy of endovascular approaches for moyamoya likely takes root in fundamental pathophysiologic differences of the disease [5, 22]. In particular, the fact that affected vessels do not show arteriosclerotic or inflammatory changes but rather smooth muscle cell hyperplasia and luminal thrombosis [5, 22] may in part explain a lack of efficacy of both angioplasty and stenting, with high rates of recurrence.

This study is limited by its very small size, brief reported follow-up periods, retrospective design, and incorporation of all reports, including case reports, which subjects results to a considerable degree of publication bias. Both cases series were from North American centers [13, 18], potentially limiting the external validity of our results to other patient populations. Nevertheless, this report strongly suggests against attempted endovascular approaches in the management of moyamoya, as both angioplasty alone and stenting are ostensibly fraught with high angiographic and clinical recurrence rates with procedural success rates similar to the combined procedural complication and initial failure rate. Although unproven, it would also seem intuitive that these approaches may prove detrimental in adult patients that may ultimately suffer from hemorrhagic moyamoya. Specifically, dual antiplatelet dependence would pose a significant problem in the context of a hemorrhagic event occurring as a result of the natural history of the disease. Although surgical revascularization is of unproven benefit for hemorrhagic moyamoya [17], it has demonstrated considerable promise in the prevention of recurrent ischemic events and should thus remain the treatment of choice.

Conclusion

An amalgamation of reports to date, although limited in number, demonstrates that endovascular treatment of symptomatic moyamoya via angioplasty and/or stenting is fraught with relatively high complication and recurrence rates. This early evidence should suggest against attempted endovascular management of symptomatic moyamoya.