Moyamoya disease: Experience with direct and indirect revascularization in 70 patients from a nonendemic region
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.178046
Source of Support: None, Conflict of Interest: None
Introduction: Moyamoya disease (MMD) is a chronic cerebrovascular disease characterized by stenosis/occlusion of bilateral terminal internal carotid arteries with the development of collaterals at base of the brain. We describe our experience of treating 70 patients (133 involved hemispheres, 108 hemispheres operated) surgically.
Keywords: Encephalo-duro-arterio-myo-synangiosis; modified Rankin score; moyamoya disease; superficial temporal artery-middle cerebral artery bypass
Moyamoya disease (MMD) is characterized by idiopathic, progressive arterial stenosis or occlusion of bilateral intracranial carotid arteries (ICA) and the development of extensive collateral vessels.  Gradual reduction of cerebral blood flow stimulates collateral vessel formation, producing the characteristic "moyamoya" blush seen on angiography.  It causes approximately 6% of childhood strokes.  First described in 1957 as "hypoplasia of bilateral internal carotid arteries,", the descriptive title of "moyamoya" was coined more than a decade later by Suzuki and Takaku. 
Though MMD occurs in a worldwide distribution, it is more common in Japan and Korea, with the incidence estimates of 0.35-0.54/100,000 population.  The classical presentations of MMD include transient ischemic attacks (TIAs), ischemic strokes, and intracranial hemorrhages. Its natural history is often progressive and includes recurrent ischemic episodes with neurological and cognitive deterioration. Intracranial hemorrhage is not only common in adults but has also been described in children.  Asymptomatic MMD may progress to cause ischemic or hemorrhagic strokes, , and in symptomatic patients, there is a high incidence of stroke in medically treated patients. 
An extracranial-to-intracranial revascularization is the main aim of surgical intervention to augment blood flow of the affected hemispheres. Direct revascularization includes anastomotic bypass between the superficial temporal artery (STA) and middle cerebral artery (MCA), whereas multiple other techniques are used for indirect bypass procedures including encephalo-duro-arterio-myo-synangiosis (EDAMS). Most clinical series have been reported from Japan and Korea; there are studies from Europe and the US as well. However, there is sparse data on moyamoya disease from the Indian population. We reviewed our experience of treating MMD at a nonendemic region and discuss the surgical strategy and outcome.
The case files of all patients who underwent surgical treatment for MMD at our institution from 2006 to 2014 were reviewed. The demographic and disease-specific data at presentation, including the type of symptoms, the baseline modified Rankin scale (mRS) score, the hemisphere requiring treatment, and the type of treatment were recorded. The imaging data were reviewed.
Only patients with either symptoms related to MMD or who had perfusion deficits along with clear angiographic evidence of MMD were selected for surgical intervention. Our strong stand towards surgical intervention is validated by the published literature, which advocates surgery to offer protection from future strokes, hemorrhages, and cognitive decline.  Since the past 5 years, the primary intent was to perform combined revascularization, which includes STA-MCA bypass along with indirect revascularization, i.e. EDAMS, for all patients, depending on the intraoperative feasibility for perforimg the procedures, irrespective of the patient's age. In children, if a recipient vessel of appropriate diameter was not available (<0.7 mm) during surgery, an indirect bypass (EDAMS) was performed.
Under general anesthesia, the patient was positioned supine with neck turned to the opposite side. STA was harvested with the standard technique, proximal to distal. A small craniotomy over the distal portion of the Sylvian fissure was done to facilitate an end-to-side anastomoses of the parietal STA branch to a cortical branch of the MCA using interrupted 10-0 nylon sutures. The mean arterial blood pressure was maintained at 0-10 mmHg above the baseline, and barbiturates were administered during the MCA cross-clamp time. The patency of the anastomosis was assessed intraoperatively with a micro-Doppler or indocyanine green angiography. An additional dural-myosynangiosis was done in a standard fashion. In the cases with a fragile STA or MCA, or with cortical branches <0.7 mm, the indirect revascularization procedure of EDAMS was preferred.
After reviewing the patient case files and the recent follow-up assessment, the mRS scores were retrieved at baseline, discharge, and follow-up. The previously published moyamoya studies that have included a large number of patients have also used the mRS score for evaluation. ,, For the purpose of our review, the patients were compared and studied in distinct pediatric and adult groups, with the combined and solely the indirect revascularization procedures being evalated separately. Patients who underwent a combined anastomosis on one side and an indirect procedure on the contralateral side (n = 10) were excluded from the clinical outcome analysis.
The primary end points in the study were the incidence of recurrent TIA or stroke in the follow up period, and the clinical outcome as measured by mRS at follow-up. Presence of a new stroke was defined as new or worsened neurological symptoms (without further improvement), along with new imaging findings consistent with the occurrence of a fresh stroke. Imaging follow-up was done using magnetic resonance imaging (MRI) or postoperative angiography, depending on the availability of these investigations. Imaging assessment was done independently by 2 observers. As many patients were from different states, uniformity in the conduction of postoperative imaging could not be maintained. The postoperative angiographic revascularization was evaluated by Matsushima and Inaba grading,  where the revascularization was graded as excellent (A, more than 2/3 rd ), good (B, 1/3-2/3 rd ), and fair (C, <1/3 rd ), based upon the extent of MCA territory supplied by external carotid artery [ECA] injection.
Fisher's exact test was used for comparing the categorical variables when the expected cell frequencies were <5, and chi-square test was used if the cell frequencies were >5. For continuous variables, comparison of means between groups was performed with analysis of variance. A Bonferroni correction was used for post hoc comparison. An independent sample t-test was used to compare the difference in the mean values between the two groups. Statistical significance was fixed at 5%. Kaplan-Meier survival analysis was used to estimate the survival function for the stroke-free subjects with the time of stroke as the end point. Log-rank (Mantel-Cox) test was used for comparing the survival distributions of the two groups. When no stroke had occurred during the follow-up period, the time listed was the last available clinical or radiographic follow-up (i.e., after this point, the occurrence or absence of stroke was censored, as is customary for survival curve analysis). All statistics were performed with SPSS version 20 (IBM, Chicago).
A total of 70 patients (mean age - 13.8 years, median - 8 years, range 2-46 years) were treated in this time period. Sixteen patients were adults (≥18 years) and 56 were in the pediatric age group. There were 34 males and 36 female patients with seven patients having a unilateral disease, out of which 4 were in the pediatric age group. A total of 133 hemispheres with the disease were evaluated out of which 108 hemispheres were surgically treated (58 combined revascularizations and 50 EDAMS) [Table 1] and [Figure 1]. In children, 23 out of 29 hemispheres with disease were operated, where 17 and 6 hemispheres were treated with combined and indirect anastomosis, respectively. In adults, 85 of the 104 hemispheres with the disease were operated, where 41 and 44 hemispheres were treated with combined and indirect anastomosis, respectively.
The most common presentation of MMD was due to symptoms related to cerebral ischemia, including both TIAs and prior infarction [Table 2] and [Figure 2]. Ninety percent of the patients presented with ischemic symptoms and only 10% had intracranial bleeding [including 2 patients with subarachnoid hemorrhage (SAH)]. In adults, only 25% patients had a hemorrhagic presentation compared to 75% with ischemic ones. Although the proportion of patients having hemorrhagic MMD was higher in adults as compared to children, 75% of the adult patients had presented with ischemia.
Out of the 70 cases, MRI was available for review in 65 cases. A border zone infarct was the most common manifestation (50% overall, 34.6% in adults, 53.8% in kids). Gyral pattern infarcts was the second most common presentation in adults (23%) and children (30%). MRI was normal in 27.6% of hemispheres [Table 3] and [Figure 3]. Angiography was the gold standard for confirmation of diagnosis and was done in all the patients in the preoperative period. Majority of the hemispheres (76%) were in Suzuki Grade III disease. All, except for 2 hemispheres, in adult patients, were in Suzuki Grade III, and 74/103 pediatric hemispheres (72%) were in Suzuki Grade III disease [Table 4]a.
The mean mRS at presentation, discharge, and follow-up were 2.27 ± 1.034, 2.46 ± 1.125, and 1.80 ± 1.269, respectively [Table 5] and [Figure 4]. The mean mRS scores of patients were analyzed at admission, discharge, and at follow-up (mean 15.9 months) by estimated marginal means of measure with Bonferroni correction. There was a significant improvement in the overall clinical status of the patients from the admission to the follow-up period (P < 0.001). The clinical improvement (from admission to follow-up) was better in pediatric patients as compared to adults (P < 0.001 vs. 0.769). When the type of surgery was compared by a multivariate analysis, patients who underwent combined revascularization (either a single side or both sides) had better clinical improvement than those patients who underwent indirect anastomosis (either only one side or both sides) [P = 0.024 vs. 0.0312].
Fifteen postoperative clinical events in 12 patients were documented, out of which 9 events occurred children from 1-3 years after surgery. Nine events occurred after the indirect anastomosis, and 6 after the combined procedure. Twelve patients developed self-limiting neurological deficits in the postoperative period, which were probably due to hyperperfusion. Computed tomography scan in all these 12 patients did not show any fresh infarcts, and the postoperative deficit (mostly speech abnormality) settled in 3-4 days on conservative treatment. Rest of the 3 events were postoperative strokes resulting in permanent morbidity [Table 6]. There was one mortality in a pediatric patient who had infection, hyperpyrexia, and then suffered a stroke after surgery performed on the second side, which resulted in death. The remaining 66 patients had a good outcome at the last follow-up. The morbidity was 2.8% per hemisphere whereas the mortality was 1.4%.
Postoperative angiograms were done to assess 44 operated hemispheres in 27 patients after a period ranging from 5 to 18 months (median 9 months) after surgery [Figure 5] and [Figure 6]. The filling of MCA vessels through the ECA injection was graded by Matsushima and Inaba grading.  The proportion of patients having A and B angiographic outcomes was higher in patients undergoing the combined revascularization, especially in the adult subgroup. However, these differences were not statistically significant. The angiographic outcome did not statistically correlate with the preoperative angiographic Suzuki grading, age, or sex [Table 4]b.
We present a large cohort of MMD patients in a nonendemic region in an attempt to extend our knowledge of the disease prevailing in India. The epidemiological studies on MMD are heterogeneous with nationwide surveys available only from Japan  and Korea.  It has been reported that the annual incidence of MMD in California and Washington states is 4-16 times lower,  and in Europe,  to be 1/10 th of the incidence of the disease in Japan. There are only sporadic case reports, and only few institutional experiences reported in literature from India, ,,,, and the exact incidence of MMD in India is unknown.
Two peaks for the age distribution of MMD have been demonstrated, the first between 45 and 49 years, and the second between 5 and 9 years. , In this study, the peak age for adult patients was 33.2 years, which is lower than that of the patients in Japan and Korea.  For children, the peak age was 7.9 years, which is also lower than reported from the latter two countries.  Women have a higher incidence than men, with the female/male predominance ranging from 1.8:1 to 4.25:1. ,,,, In our study, the sex ratio (men to women) of the patients was 1.05, which is lesser than that found in Japan and Korea , but is comparable to that oberved in studies from China.  There were no familial forms of MMD in this series compared to 5.6-15% reported in the Japanese population. 
In this study, the presenting symptoms were predominantly ischemic in both adult and pediatric subjects [Table 2] in contrast to the patients in China, Japan, and South Korea ,,, but comparable to studies from USA and Europe, where adults seem to present with a lower rate of hemorrhage than in East Asian countries. , Even though the role of surgical revascularization in hemorrhagic MMD is still controversial, a recent randomized trial  as well as other studies  prove the benefit of direct revascularization in prevention of future hemorrhages in these patients.
Intracranial aneurysms are found in 3-15% of patients with MMD.  In our series, two patients had aneurysms which bled, with the patients presenting with subarachnoid hemorrhage. One adult female patient had a SAH from right superior cerebellar artery aneurysm, whereas another boy had a right supraclinoid ICA aneurysm. Both aneurysms were surgically clipped, and EDAMS was performed in the same sitting. For MMD associated with intracranial aneurysms, an urgent treatment should be performed especially if the aneurysm has bled. The aneurysm can be managed with either surgical clipping or endovascular coiling.
Cho et al.,  studied the infarct patterns in MMD and found that gyral pattern of infarction was the most common form (44.0%). Border zone and deep lacunar patterns were infrequent. Gyral and border-zone patterns were more frequently seen in childhood, whereas a honeycomb pattern was not seen in young patients. In this series, border-zone infarcts followed by gyral infarcts were most common in adults, while border zone infarcts followed by territorial infarcts were common in the pediatric age group [Table 3] and [Figure 3]. This is a finding similar to that found in a recent report by Rafay et al. 
Direct revascularization by STA-MCA bypass provides the benefit of both immediate and progressive revascularization and the anastomoses are proven to enlarge up to 50% with time.  Some disadvantages of the direct anastomoses are the precipitation of stroke from the embolic event, cross clamping of recipient vessels, the occurrence of the hyperperfusion syndrome, a prolonged operative time as well as the technical challenges of anastomosing arteries having a submillimeter diameter. With experience and skill development, the operative time can be shortened and careful case selection can avoid complications. Guzman et al.,  in their series of 450 revascularization procedures in children and adults, reported that 91% patients were treated with a direct anastomoses, with the risk of stroke or hemorrhage being only 5.5% over 5 years, and 71% patients had a functional improvement. They reported that symptoms resolved within a month of the direct bypass. Other authors have also reported excellent results with direct and indirect revascularizations in patients with MMD. , Lee et al.,  showed in their study, that the extent of revascularization and decrease in the moyamoya vessels in adult patients was significantly more in patients treated with a direct bypass or a combined procedure when compared to the indirect methods (150 procedures in 106 patients).
Indirect revascularization generally needs a shorter operative time and presents with fewer technical challenges. The main disadvantage is the delay of weeks to months for new vessel formation. , Some series demonstrate an increased risk of stroke compared to the direct revascularisation. , EDAMS was the surgical procedure of choice in the indirect bypass contemplated in our institution. About 90-95% of event-free survival has been reported with the indirect bypass techniques. ,
In the present study and in the literature, functional outcomes in adults undergoing a direct bypass were significantly better than those of patients undergoing an indirect bypass.  However, there are little differences in the pediatric population in the long-term outcome with these two procedures. Studies have shown that angiographic revascularization patterns do not correlate well with clinical outcomes.  A meta-analysis of 57 publications detailing 1156 pediatric moyamoya patients who underwent 1448 revascularization procedures (25% direct, 75% indirect) reported that 87% of those undergoing revascularization had a symptomatic improvement. Indirect, direct, or combined revascularization procedures were all equally effective. 
Literature from India is sparse and restricted to a few case series. Garg et al.,  analyzed 44 patients (18 pediatric) out of whom, 9 had undergone surgery. They showed that conservative treatment led to recurrent symptoms in 36.8% of their patients whereas operated patients were symptom-free. Sundaram et al., reported their study of 36 patients (15 pediatric) where three surgically treated patients had recurrent ischemic events, but none of the conservatively treated patients had these events. Controversially, clinical outcome was better with the conservatively managed patients. Recently, Sahoo et al.,  reported their experience with 33 operated MMD patients (36 EDAMS and 4 EDAMS + STA−MCA bypass), out of which 17 were pediatric (21 treated hemispheres), and 16 were adult patients (19 treated hemispheres). They proposed a new angiographic scoring system based on the reformation of distal MCA and anterior cerebral artery, regression of basal moyamoya vessels, leptomeningeal collaterals, as well as the overall perfusion. Matushima Grade A perfusion was seen in 73% patients in the pediatric and in 44% of the adult hemispheres, whereas Matushima Grade B perfusion was seen in 18% of the pediatric and 44% of the adult hemispheres, but the differences were not statistically significant. Apart from these reports on operative treatment, various case series and institutional reviews of nonoperative management ,, and anesthetic management ,, have been reported.
In this series, we report the largest series of patients with MMD from India. We were able to compare a relatively large group of both pediatric and adult patients who were treated by either combined or indirect bypass procedures. Overall, the pediatric patients fared better after surgery with a significant clinical improvement compared to adults. When individual surgeries were compared, both combined and indirect revascularization procedures were associated with significant clinical improvement. However, the combined revascularization procedure was associated with better improvement compared with solely the indirect revascularization surgery. Angiographic outcome did not correlate with age, sex, the preoperative Suzuki grading or even the type of surgery. However, there was a trend toward better angiographic outcome with combined revascularization. There were 3 strokes and one mortality, with an overall morbidity of 2.8%, and a mortality of 1.4%, which is comparable with the published stroke rates after surgery. , There was no significant difference in the stroke rates between the combined and the indirect revascularization surgeries.
Though a prospective randomized controlled clinical trial is necessary to establish a definitive conclusion about the efficacy of direct and indirect bypasses, it is not ethically feasible, as an intraoperative randomization will have to be done in cases where both procedures are feasible. In this series, despite equivalent protection from stroke, outcomes were better after the combined revascularization. As combined revascularization gives better results, and there is no increase in the risk of complications with the former procedure as compared to the indirect revascularization, we recommend a combined revascularization for MMD, wherever feasible.
This series demonstrates that both combined and indirect revascularization procedures are effective in the treatment of MMD in preventing strokes in both adult and pediatric populations. Our pediatric patients had a better clinical improvement after surgery than the adults. We, therefore, believe that combined revascularization surgery should be the surgical strategy in all age groups as it is feasible in a significant group of pediatric patients too.
We would like to acknowledge Tharangini V, Ph.D, for her help in statistical methods and analysis.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]