Transcranial Doppler in Childhood Moyamoya: an Underutilized Tool?

Abstract

INTRODUCTION:

Children with moyamoya are at high risk for incident and recurrent stroke. Transcranial Doppler (TCD) ultrasound is an attractive option to screen high-risk populations for moyamoya and to provide stroke risk stratification information due to its safety and cost effectiveness. We used TCD to evaluate cerebral blood flow velocities in children with presurgical moyamoya and to determine if velocities differ between children with stable versus unstable disease.

METHOD:

Fourteen participants ≤ 21 years old with a radiographic diagnosis of moyamoya or moyamoya-like arteriopathy underwent a research TCD at a median age of 7.2 years. TCDs were performed outside of the setting of acute stroke and prior to surgical revascularization. Arteriopathy was classified as unstable if the participant had a stroke or transient ischemic attack within 3 months preceding the TCD.

RESULTS:

Middle cerebral artery and internal carotid artery blood flow velocities were elevated. The median M1 velocity was 138 cm/s (IQR 106-168). Individual M1 flow velocities were a median of 5.0 standard deviations above age-based normative values. The median distal ICA velocity was 146 cm/s (IQR 124-163). Individual ICA flow velocities were a median of 5.9 standard deviations above normative values. Participants with unstable arteriopathy had higher M1 velocities compared to those with stable arteriopathy (170 versus 119 cm/s, p=0.0003). We did not identify velocity differences based on comorbid conditions or age.

CONCLUSIONS:

These preliminary data suggest that TCD is a promising tool for screening for cerebral arteriopathies in high-risk pediatric populations and assessment for unstable disease.

Moyamoya arteriopathy, defined by progressive stenosis of the intracranial arteries in a characteristic pattern with development of collateral vasculature, accounts for 6-10% of childhood strokes and transient ischemic attacks (TIAs) and is associated with high recurrent stroke risk and significant morbidity.^1-4 Moyamoya syndrome refers to moyamoya arteriopathy associated with a comorbid condition such as sickle cell disease (SCD), Down syndrome, neurofibromatosis, or history of cranial radiation therapy, while moyamoya disease refers to the arteriopathy in isolation.^5,6

Early diagnosis of moyamoya and surveillance for unstable disease, defined as a situation in which cerebral metabolic demand exceeds perfusion capacity, are important for decisions regarding candidacy for and timing of surgical revascularization. Children with SCD are screened for stroke risk with transcranial Doppler (TCD) ultrasonography starting at age 2 years, but no evidence-based screening recommendations exist for children with other high-risk conditions, such as Down syndrome and after cranial radiation therapy. TCD is an attractive option to screen for moyamoya and to identify unstable disease because it provides hemodynamic information and does not require anesthesia, intravenous lines, or contrast/tracer agents.⁷ TCD can be performed at the bedside, at low-cost, and in resource-limited settings.

We evaluated baseline TCD values for children with moyamoya and moyamoya-like arteriopathies prior to surgical revascularization, and in an exploratory analysis, aimed to determine if TCD parameters can distinguish between stable and unstable disease.

Methods

In this prospective observational study, consecutive patients ≤21 years old with a radiographic diagnosis of moyamoya or moyamoya-like arteriopathy treated at a single tertiary care center between June 2019 and December 2022 were approached about study participation. Participants underwent a research TCD outside of the setting of acute stroke. Patients who had prior surgical revascularization were excluded. The Johns Hopkins Medicine Institutional Review Board approved the study (IRB00252581), and all families signed informed consent.

Demographic, clinical, and radiographic data were collected. Participants were classified as having unstable arteriopathy if the participant had an ischemic or hemorrhagic stroke or TIA within the 3 months preceding the TCD. Classic moyamoya was defined as steno-occlusive disease of the bilateral carotid termini with lenticulostriate collateral vascular development, while moyamoya-like arteriopathy was defined as having some but not all features of moyamoya. Individual vessels were characterized as diseased if there was evidence of stenosis or occlusion on magnetic resonance angiography (MRA).

All TCDs were performed while children were awake and calm in a sitting or supine reclined position. TCDs were performed by two experienced vascular sonographers using a DWL Doppler (Germany) non-imaging machine. Mean cerebral blood flow velocities (CBFVs) at the M1 segment of the middle cerebral artery (MCA) and the distal internal carotid artery (ICA) were compared to age-based normative values.⁸ Blood pressure, temperature, and heart rate were measured at the time of the TCD. Hypertension was defined as systolic or diastolic blood pressure ≥95^th percentile for age, sex, and height.⁹ Study data were managed using research electronic data capture (REDCap) tools hosted at Johns Hopkins University.¹⁰

Statistical analysis was performed using Stata 18.0 (StataCorp, College Station, TX). Medians and interquartile ranges were calculated for demographic data and blood flow velocities. Differences in CBFVs between groups were assessed using Wilcoxon rank sum tests.

Results

Of 29 potential participants meeting eligibility criteria, 14 participants were enrolled and underwent TCDs at a median age of 7.2 years (range 3.8-21.4 years) (Table). We identified no demographic or disease-related differences between the 14 participants and the 15 eligible nonparticipants (Supplementary Table 1). Three participants (21%) had moyamoya disease and 11 (79%) had moyamoya syndrome. Six participants (43%) had an ischemic stroke and one (7%) had a hemorrhagic stroke prior to enrollment. No participant had revascularization surgery prior to the TCD. At the time of the research TCD, arteriopathy was classic in eight (57%), moyamoya-like in six (43%), and unilateral in five participants (36%). The arteriopathy was classified as unstable in six (43%), with stroke or TIA occurring a median of 5 days prior to the research TCD in those cases.

Table.

Demographic, Clinical, and Radiographic Participant Information

		Whole Cohort (n=14)	Participants without SCD (n=8)
Sex, n (%)	Male	8 (57%)	4 (50%)
Age at moyamoya diagnosis, median (range)		6.8 years (3.4-21.3 years)	7.6 years (3.4-21.3 years)
Age at TCD, median (range)		7.2 years (3.8-21.4 years)	7.9 years (3.8-21.4 years)
Hemoglobin within 14 days of TCD, median (IQR) (n=9)		10.6 g/dL (10.0, 12.5)	11.6 g/dL (10.2, 13.0)
Associated Condition, n (%)	None (moyamoya disease)	3 (21%)	3 (38%)
	Sickle cell disease	6 (43%)	N/A
	Neurofibromatosis-type 1	1 (7%)	1 (13%)
	Hemophilia A	1 (7%)	1 (13%)
	Genetic systemic disease	2 (14%)	2 (25%)
	Morning glory disc anomaly	1 (7%)	1 (13%)
Arteriopathy Laterality at Diagnosis, n (%)	Bilateral / symmetric	4 (29%)	3 (38%)
	Left > Right	5 (36%)	3 (38%)
	Right > Left	0	0
	Left only	2 (14%)	1 (13%)
	Right only	3 (21%)	1 (13%)
Arteriopathy Classification	Classic moyamoya	8 (57%)	5 (63%)
	Moyamoya-like	6 (43%)	3 (38%)
Arteriopathy Stability	Unstable	6 (43%)	4 (50%)
	Stable	8 (57%)	4 (50%)
History of Clinical Ischemic Stroke, n (%)		6 (43%)	3 (38%)
History of Clinical Hemorrhagic Stroke, n (%)		1 (7%)	1 (13%)

TCD, transcranial Doppler

Ten participants (71%) were on antiplatelet medications at the time of the TCD. Reasons participants were not on antiplatelet therapy included sickle cell disease (n=2), hemophilia A (n=1), and intracranial hemorrhage (n=1). Four participants (29%) were on antiseizure medications (lacosamide and/or levetiracetam), five (36%) with SCD were on chronic transfusion therapy, and one (7%) with SCD was on hydroxyurea. No child was febrile or ill during the study. Seven participants (50%) were hypertensive for age. One participant (7%) was tachycardic. Nine children had blood counts performed within 14 days of the TCD; median hemoglobin was 10.6 g/dL (lower limit of normal: 12.4 g/dL). One TCD was aborted early due to patient agitation; the remainder had adequate bone windows and sufficient technical acquisition of data. No adverse events were associated with TCD performance.

Compared to age-based normative values, MCA and ICA CBFVs were elevated. The median M1 CBFV was 138 cm/s (IQR 106-168); individual CBFVs were a median of 5.0 standard deviations above age-based norms (IQR 3.2-8.7). Median distal ICA CBFV was 146 cm/s (IQR 124-163); individual CBFVs were a median of 5.9 standard deviations above age-based norms (IQR 3.1-8.3).

Participants with unstable arteriopathy had higher M1 CBFVs compared to those with stable arteriopathy (170 v 119 cm/s, p=0.0003) (Figure, Supplementary Table 2). Participants with classic moyamoya had higher M1 CBFVs compared with those with moyamoya-like arteriopathies (159 v 129 cm/s, p=0.042). There was no difference between CBFV of diseased MCAs (n=21) versus MCAs without steno-occlusive disease (n=5) (145 v 132 cm/s, p=0.63). There were no differences in M1 CBFVs based on the presence of a comorbid condition (p=0.19), presence of sickle cell disease (p=0.56), or age (<7 v >7 years, p=0.13).

Figure.

Distribution of MCA (left) and ICA (right) cerebral blood flow velocities in participants with unstable (black circles) and stable (open circles) cerebral arteriopathy.

Discussion

In this cohort of 14 children with presurgical moyamoya and moyamoya-like arteriopathies, CBFVs measured by TCD were elevated compared with age-based normative values. Participants with unstable disease had significantly higher CBFVs compared with those with stable disease. These exploratory data suggest a role of TCDs in identifying and risk-stratifying childhood arteriopathies.

The mechanism of increased CBFVs in these children is unclear. CBFV can be elevated due to distal focal vascular narrowing.⁷ Our data argue against this as the sole mechanism, as the velocities did not significantly differ between vessels with and without steno-occlusive disease. This is concordant with published data showing that the minority of children with SCD and elevated CBFVs on TCD have intracranial stenosis on MRA.^11,12 Abnormal velocities may therefore represent global aberrant flow dynamics and contribution of collateral flow. Hyperdynamic flow may also increase CBFVs. Some children in this cohort were hypertensive, presumably in hemodynamic compensation for cerebral arteriopathy, and some were anemic, with a median hemoglobin of 10.6 g/dL.

Currently, the optimal treatment approach for children diagnosed with moyamoya before they develop ischemic symptoms is unknown.¹³ Identification of unstable disease before it is symptomatic may improve patient selection for surgical revascularization. Whether TCD can identify the point at which the disease becomes unstable (when cerebral metabolic demand exceeds perfusion) is unknown and warrants investigation in long-term follow up studies.

This study is exploratory and has important limitations. Sample size is small; findings require confirmation in larger cohorts. We included both the left and right hemisphere CBFVs for each participant in our statistical analysis, even though the two hemispheres are not independent of each other given connection through the Circle of Willis. Though a limitation, this practice is concordant with prior studies that include both hemispheres as if they were independent samples. The sample was too small to account for arteriopathy severity, but future studies should evaluate how the degree of stenosis impacts CBFVs. Finally, arteriopathy was defined as unstable if the participant had a stroke, TIA, or other new symptom attributable to moyamoya within 3 months preceding the TCD, presumably reflecting that cerebral metabolic demand exceeded perfusion capacity. However, this assumption was not confirmed using perfusion imaging, and cerebral hemodynamic instability causing stroke may be transient and related to dehydration, anemia, or other factors; thus, it is possible that disease was not truly unstable at the time of the TCD.

Future studies should focus on screening high-risk populations and determining the sensitivity and specificity of different CBFV thresholds in identifying cerebral arteriopathies. Future research should also evaluate if TCD can predict stroke in children with moyamoya. If so, TCD may be a valuable tool to assist in selection for revascularization surgery, especially in pre-symptomatic patients. Finally, TCD may have a role in predicting non-stroke outcomes. For example, poor neuropsychological outcomes are common in children with moyamoya, even in the absence of clinical stroke, and improvement in cerebral hemodynamic status after surgery has been associated with neuropsychological improvement.^14,15 Whether TCD can predict neuropsychological decline should be explored in future investigation.

Conclusion

TCDs is a promising tool for screening for cerebral arteriopathies in high-risk pediatric populations and assessment for unstable disease that may benefit from surgical revascularization. Future studies should evaluate if TCD can predict future stroke in children with moyamoya and determine the sensitivity and specificity of different CBFV thresholds in screening high-risk populations for cerebral arteriopathies.