Late Cerebrovascular Events and Social Outcome after Adolescence: Long-term Outcome of Pediatric Moyamoya Disease

Abstract

In this article, the authors review the literature related to long-term outcome of pediatric moyamoya disease, focusing on late cerebrovascular events and social outcome of pediatric patients once they reach adulthood. Late-onset de novo hemorrhage is rare but more serious than recurrence of ischemic stroke. Long-term follow-up data on Asian populations suggest that the incidence of de novo hemorrhage might increase at age 20 or later, even more than 10 years after bypass surgery. Social adaptation difficulty, possibly related to cognitive impairment caused by frontal ischemia, continues in 10–20% of patients after they reach adulthood, even if no significant disability is present in daily life. A treatment strategy aimed at improving long-term outcome and careful follow-up might be required.

Introduction: Long-term Outcome of Pediatric Moyamoya Disease: It’s Time to Investigate

Do children with moyamoya disease enjoy a satisfactory social life once they reach adulthood? Are they able to study and work as well as their healthier counterparts? Do they enjoy long lives free from concerns about stroke after bypass surgery? All these questions are relevant to parents of a child with moyamoya disease. However, these questions, also important to clinicians,¹⁾ which have remained unanswered.

The time has come for researchers to address the issues of social outcome and late cerebrovascular events after adolescence in pediatric moyamoya disease. The first successful surgical treatment for moyamoya disease – superficial temporal artery (STA) to middle cerebral artery (MCA) anastomosis, also described as “direct bypass” – was reported in late 1970s by Karasawa et al.²⁾ In the early 1980s, Matsushima et al.³⁾ introduced an indirect bypass known as encephalo-duro-arterio-synangiosis (EDAS). It is only since the 1980s, therefore, that surgical treatment has commonly been applied to the disease. Many children who had undergone surgery at that time have been entering adulthood.

Indeed, some recent studies have accumulated more than a decade of follow-up data.^4–8) Such extremely long-term data might provide important clues to solve the above questions. Accumulation of reliable data is not always easy, however, because of the heterogeneity of surgical procedures and loss of follow-up, both of which could introduce critical bias to the study results.

The purpose of the present article is to review literature on the long-term outcome of pediatric moyamoya disease, focusing on late cerebrovascular events and social outcome after patients have entered adulthood.

Late Cerebrovascular Events, Especially De Novo Hemorrhage

Bypass surgery seems to prevent ischemic symptoms, the most common symptoms in pediatric patients, for years. Follow-up studies in pediatric moyamoya disease demonstrate the eventual elimination of transient ischemic attacks (TIA) after surgery in 93–98% of patients.^7,9–13) Recurrence of ischemic stroke is also rare; only a small proportion (0–6.3%) of patients experience late ischemic stroke (Table 1).^{4,6,8–10,14–18)} In the authors’ data, only one case experienced ischemic stroke 2 years after surgery, the event occurring after craniotomy for a severe head injury from a traffic accident.⁸⁾ The incidence of late-onset ischemic stroke in our data was as low as 0.10% per year (Fig. 1).

Table 1.

Incidence of late cerebrovascular events^* after surgery in the literature of pediatric moyamoya disease, stratified by population size and length of follow-up

Size†	F/u length‡	Authors	Treatment	No. of cases	Mean F/u period (years)	F/u rate (%)	No. of late ischemic strokes (%)	No. of late hemorrhagic strokes (%)	Overall incidence of stroke (% per year)
Large	Extremely-long	Mukawa et al.6)	Indirect	172	14.3	83	3 (1.7)	3 (1.7)	0.24
		Funaki et al.8)§	Direct (combined)	58	18.1	96.6	1 (1.7)	3 (5.2)	0.41
		Imaizumi et al.4)	Various	25	18.8	80.6	1 (4.0)	3 (12.0)	0.85
Large	Long	Bao et al.9)	Indirect	288	4.4	N/A	8 (2.8)	2 (0.7)	9.00‖
		Scott et al.16)	Indirect	126	5.1	99.2	4 (3.2)	0	NA
		Kuroda et al.10)	Direct (combined)	28	6.1	N/A	0	0	0
		Rashad et al.14)	Direct (combined)	23	6.4	95.7	0	0	0
Small	Extremely-long	Isono et al.18)	Indirect	11	12.8	N/A	0	0	0
		Goda et al.17)	Indirect	6	15.2	N/A	0	0	0
Small	Long	Darvish et al.15)	Various	16	7.3	N/A	1 (6.3)	0	N/A

^*Excludes perioperative strokes,

^†“Large” indicates number of patients exceeds 20,

^‡“Extremely long” indicates a mean F/u period exceeding 10 years,

^§The authors’ group,

^‖Includes perioperative strokes.

Fig. 1.

Kaplan–Meier Curves showing incidence of late-onset ischemic stroke (left) and hemorrhage (right) following direct bypass surgery for pediatric moyamoya disease in the authors’ group.

De novo intracranial hemorrhage, a rather unique transformation from ischemic to hemorrhagic moyamoya disease, might be the more serious issue in adulthood, according to data on Asian populations.^4,6,8) Hemorrhage is the factor most severely affecting outcomes in moyamoya disease and is often fatal.¹⁹⁾ The incidence of such de novo hemorrhage is estimated at 0.30% per year after surgery according to the authors’ data, higher than that of late ischemic stroke (Fig. 1).⁷⁾ The incidence is quite low, however, compared with the de novo hemorrhage rate in asymptomatic adult patients (2–3% per year);^20,21) hence, surgery in childhood is probably beneficial for primary prevention of hemorrhage. Yet the risk of de novo intracranial hemorrhage cannot be underestimated, given that it increases with age.

The data support this hypothesis. As shown in Table 1, late onset hemorrhagic events were reported in all three extremely long-term follow-up studies, the follow-up period of which exceeds 10 years.^4,6,8) Two of these studies described the initial presentation at disease onset,^4,8) and all late-onset hemorrhages included in the two studies occurred in the patients who had not initially presented with intracranial hemorrhage at disease onset. We calculated the frequency of late cerebrovascular events using data pooled from previous studies^{4,6,8–10,14–18)} and stratified by the length of study period (Fig. 2). More hemorrhages are observed in studies exceeding 10 years than in those of less than 10 years. In the authors’ data, all de novo hemorrhages occurred more than 10 years after bypass surgery (Fig. 1). This finding is similar to other extremely long-term follow-up studies.^4,6) De novo hemorrhage tends to occur at age 20 or later;⁸⁾ in authors’ series, all hemorrhages occurred in patients in their 20s and 30s. This is consistent with epidemiological findings that hemorrhagic presentation in the disease increases at age 20.²²⁾

Fig. 2.

Graph showing frequency of late cerebrovascular events (ischemic stroke and hemorrhage) in pediatric moyamoya disease using data pooled from previous studies,^{4,6,8–10,14–18)} stratified by length of study period.

Long-term Management Focusing on De Novo Hemorrhage

The above findings suggest that the pediatric patients should be monitored carefully for more than 10 years after surgery in light of the risk of de novo hemorrhage after adulthood. Follow-up for a period of less than 10 years might underestimate the risk (Table 1).^{9,10,13–16,23)} For female patients in particular, an examination might be recommended before pregnancy considering the risk of intracranial hemorrhage during pregnancy and delivery, although most patients diagnosed with moyamoya disease are able to give birth safely when carefully managed.²⁴⁾

Antiplatelet agents are often administrated in medical treatment and perioperative management of pediatric moyamoya disease.^14,22,25) These agents should be discontinued when possible before adulthood in light of their adverse effect on hemorrhage. In the authors’ data, administration of antiplatelet agents was discontinued after surgery in 92.7% of pediatric patients before adulthood. Risk factor control, especially blood pressure management, is also important in preventing de novo hemorrhage. Hypertension can occur even in young patients because renal artery stenosis can accompany moyamoya disease.²⁶⁾ Lifestyle advice, including efforts to discourage smoking in adulthood, might also be advisable to prevent de novo hemorrhage.

The mechanism of de novo hemorrhage remains unclear. In our cases of de novo hemorrhage, collaterals from the choroidal artery or thalamic perforators had been identified before hemorrhage despite the good patency of the bypass.⁸⁾ Recent studies on hemorrhagic moyamoya disease suggest that such collaterals are at higher risk of rupture than are moyamoya vessels.²⁷⁾ Further studies focusing on angiographic risk factors of de novo hemorrhage might be required.

Social Outcome after Adolescence

While most pediatric patients are capable of independent daily activity for a long time after surgery,^6,7,14,16) recent studies have revealed that 10–20% of patients have difficulty with social adaptation after adolescence. Nakashima et al.²⁸⁾ reported that about 10% of patients had severe difficulty in social or school life because of intellectual impairment. Miyamoto et al.¹⁾ reported in their questionnaire-based study that 13.3% of patients had difficulty carrying on an independent social life because of minor intellectual delays, although the study included both pediatric and adult patients. Phi et al.⁵⁾ also conducted a long-term follow-up study with questionnaire and reported that the prevalence of pediatric patients who eventually entered college or university was 80% and that 62% of patients found employment in adulthood. They also reported, however, 20% of the responders were dissatisfied with their treatment outcomes. In the authors’ study, 17.9% of pediatric patients with moyamoya disease continued to suffer from social adaptation difficulties, such as difficulty with regular school attendance or obtaining regular employment.⁷⁾

Several risk factors associated with unfavorable social or functional outcome have been reported to date^{4,5,7,12,13,29–32)} (Table 2). Among them, preoperative neurological deficit, preoperative infarct, involvement of the posterior cerebral artery, and indirect bypass with small craniotomy have been identified as risk factors through multivariate analyses.^5,7,12) Preoperative neurological impairment and major stroke are probably the strongest factors associated with unfavorable social outcome.^{4,5,7,13,30,32)} On the other hand, a recent study demonstrates that some TIA-type pediatric patients, exhibiting no focal intracranial lesion, have specific cognitive impairments.³³⁾ Several studies on adult patients support this finding. Karzmark et al.³⁴⁾ demonstrated that executive functioning, classified as frontal lobe function, could be impaired in the absence of ischemic stroke. The study of the authors’ group also demonstrated that those suffering social adaptation difficulty with minimum brain lesion showed significantly lower scores in several frontal lobe assessment tasks.³⁵⁾ Nakagawara et al.³⁶⁾ demonstrated that significant cortical neuron loss, detected with [¹²³I]-iomazenil single photon emission computed tomography, had occurred in the frontal lobes of patients with higher brain dysfunction.

Table 2.

Risk factors associated with unfavorable social or functional outcome in pediatric moyamoya disease

Type	Factor
Unchangeable risk factors	Preoperative neurological impairment4,5,32)* Preoperative major stroke/infarct2,7,30)*
Indicator of early surgical intervention	Onset at younger age2,4) Posterior cerebral artery involvement7)* Homozygous RNF213 p.R4810K variant29)
Changeable risk factors	Longer duration after onset without surgery31) Small craniotomy12)*

^*Statistically significant in multivariate analyses.

These results reasonably lead to the hypothesis that longstanding frontal ischemia since childhood can cause impairment of cognitive function, resulting in unfavorable social outcome in adulthood. Although attractive, this hypothesis remains to be tested. The correlation between cognitive dysfunction and frontal ischemia has rarely been addressed, and it is a subject of controversy. Calviere et al.³⁷⁾ demonstrated that dysexecutive cognitive syndrome is related to the frontal cerebrovascular reserve measured with perfusion magnetic resonance imaging, although their study included only 10 adult patients. They also demonstrated that elevation of the apparent diffusion coefficient in normal-appearing frontal white matter of adult patients was associated with executive dysfunction.³⁸⁾ On the other hand, Mogensen et al.³⁹⁾ demonstrated that similar executive dysfunction was associated with a posterior, not anterior, stroke. Recent studies using resting state functional magnetic resonance imaging reveal that patients with cognitive and executive dysfunction exhibit a disrupted functional connectivity network.^40–42)

Social Outcome: How Can It Be Managed?

Elimination of known risk factors associated with unfavorable social outcome is a reasonable preventive approach, although not all factors are changeable (Table 2). Younger age at onset^4,13) – especially age under 3 years – and posterior cerebral artery involvement⁷⁾ are considered indicators of early surgical intervention, because patients with these factors are at high risk of future unfavorable social outcome.

Kuroda et al reported that direct bypass with a large craniotomy extending to the frontal area was associated with better intellectual outcome.¹²⁾ Bypass surgery targeting the frontal area supplied by the anterior cerebral artery (ACA) seems an attractive approach for improving social outcome, although the reason for this is not completely clear. Several bypass procedures targeting the ACA territory have been reported to date, and these are classified into four types: simultaneous indirect revascularization of both ACA and MCA territories;^43–45) simultaneous direct and indirect revascularization of the MCA territory and ACA territory, respectively (or vice versa);^10,46) simultaneous direct revascularization of both ACA and MCA territories;⁴⁷⁾ and additional direct revascularization of the ACA territory after conventional STA–MCA anastomosis.⁴⁸⁾ As the number of patients included in these series was relatively small, no evidence exists regarding indication and optimal procedure for the ACA territory bypass. Controversy might also arise regarding the universal strategy for targeting the ACA territory; a recent study revealed hemodynamic improvement of the ACA territory after conventional MCA territory bypass alone.⁴⁹⁾ Whether bypass for the ACA territory is beneficial to cognitive function should be validated in further studies.

For children suffering from neurological or cognitive dysfunction due to pre-existing infarct, educational support addressing the cognitive characteristics of each patient and cooperation between hospital and school staff might also be made mandatory for improving quality of life.⁵⁰⁾

Conclusion

Recent long-term follow-up studies of pediatric moyamoya disease have revealed two salient issues for patients who have reached adulthood: late-onset de novo intracranial hemorrhage and social adaptation difficulty. The risk of de novo hemorrhage might increase after age 20 or later, even more than 10 years after bypass surgery, according to data on Asian populations. Careful long-term follow-up and management focusing on hemorrhage prevention in adulthood might be required. Social adaptation difficulty, possibly related to cognitive impairment caused by frontal ischemia, continues in 10–20% of patients after they reach adulthood, even if they do not have significant disability in daily life. Early surgical intervention might be considered, especially for patients at high risk of unfavorable social outcome. Although bypass procedures targeting the ACA territory are an attractive approach for preventing unfavorable social outcome, the benefit should be validated in further studies.