White and Gray Matter Perfusion in Children with Moyamoya Angiopathy after Revascularization Surgery

Introduction: Surgical revascularization is very effective in patients with moyamoya angiopathy (MMA) and leads to improvements in cortical perfusion parameters. However, changes in white matter hemodynamics are still underestimated. To date, only a few studies have examined brain perfusion changes within deep white matter after bypass surgery in patients with MMA. Methods: Ten children with MMA were evaluated using the CT perfusion technique before and after revascularization surgery. Brain perfusion parameters within gray and white matter were compared before and after surgery. The correlations between the perfusion parameters before surgery and the Suzuki stage, as well as between the perfusion parameters and the cognitive scores, were also evaluated. Results: Brain perfusion parameters improved significantly in both gray matter (predominantly due to cerebral blood flow within the anterior circulation, p < 0.01) and white matter (predominantly due to cerebral blood volume within the semiovale centrum, p < 0.001). We revealed that the pattern of improvement in perfusion in white matter differed from the pattern of improvement in perfusion in gray matter. Significant correlations were revealed between the Suzuki stage before surgery and the perfusion parameters within the posterior cerebral artery circulation (adjusted p < 0.05). There were also significant correlations between cognitive scores and brain perfusion parameters in gray matter and white matter (adjusted p < 0.05). Conclusions: The perfusion parameters of gray matter and white matter in the brain improve differently after bypass surgery in patients with MMA. Different hemodynamics within these compartments could explain this.


Introduction
Moyamoya angiopathy (MMA) is a rare cerebral vasculopathy described by steno-occlusion of the circle of Willis arteries and the development of a collateral vessel network (moyamoya vessels) [1].In moyamoya disease, MMA is characterized by progressive steno-occlusion of the terminal segments of the internal carotid arteries and/or their proximal branches as a single manifestation, whereas in moyamoya syndrome, MMA is associated with an underlying disease [2].These changes typically result in chronic ischemia of the brain parenchyma with subsequent strokes [3].
The prevalence of MMA is highest in Asia [4] and much lower in non-Asian countries [5].Asian and European patients appear to have identical pathophysiological pathways [6].It has also been shown recently that Asian and European patients with MMA have similar angiographic characteristics, suggesting similar treatment regimens worldwide [7].
In children, MMA is the leading cause of ischemic brain injury [8].In patients with this disease, the risk of recurrent stroke reaches 20% in 13 months, 9% of which are multiple [9].It should be noted that MMA also results in impairments in cognitive function, and these impairments appear before cerebrovascular accidents due to chronic hypoperfusion [10].Patients with MMA often suffer cognitive dysfunctions such as memory, attention, and behavioral disturbances [11].In addition, previously published studies have shown that MMA may have a greater impact on cognition in children than in adults [11].
Currently, there are no effective methods to reverse progressive arteriopathy; instead, surgical revascularization is the preferred method, with branches of the external carotid artery routed to the brain directly (bypass) or indirectly (synangiosis) [8].In recent years, more preference has been given to direct and combined methods of revascularization in both children [12] and adults [13] due to the development of microsurgical techniques.
Perfusion imaging is a great tool for assessing blood flow in the brain parenchyma.In addition to conventional angiography, cerebral blood flow evaluation (CBF) is widely used in patients with MMA to determine the necessity of surgical treatment [14].As previously shown, brain perfusion parameters improve after surgery in MMA patients, which subsequently leads to clinical improvement, especially in children [15].There are multiple papers in the field that focus exclusively on gray matter perfusion alterations before and after surgical revascularization [16].On the other hand, white matter perfusion changes remain underestimated.To date, only a few studies dedicated to white matter perfusion changes in patients with MMA have been published [17].However, white matter watershed regions (such as the semiovale center) are the most prone to chronic ischemic changes in MMA [18], and it is still unclear what changes can be found in these patients after bypass surgery.Thus, the purpose of this study was to estimate cerebral hemodynamic changes in gray matter and white matter in children with MMA after revascularization surgery.

Subjects
The subjects of this retrospective study were patients diagnosed with MMA according to the diagnostic guidelines [2] who were surgically treated with combined revascularization techniques at the Federal Neurosurgical Center, Novosibirsk, from 2015 to 2021.All patients underwent a series of CT perfusion scans before and 6-12 months after the intervention.All patients were also subjected to digital subtraction angiography before and 5-7 months after the intervention.

Cognitive and Neurological Evaluation
Neurological assessment was performed and quantified with a modified Rankin Scale (mRS [19]) and Barthel index [20] before and 5-7 months after revascularization surgery.Cognitive function assessment was performed using the verbal fluency test (VFT [21]), the Ten-word test from Luria neuropsychological assessment battery [22], and the Trail Making Test (TMT [23]) for seven children before and 5-7 months after revascularization surgery.

Acquisition of CT Perfusion Data
CT perfusion data were acquired in the axial plane using a 64row multidetector CT scanner (Somatom Definition, Siemens) before and after surgical intervention for each patient.The shuttle perfusion technique was used, and the scan volume with 80 mm craniocaudal coverage was placed (with coverage of the entire hemisphere).The acquisition parameters of the scan were 80 kVp, 160-180 mAs, 8 mm slice thickness, a field of view of 22 cm, and an image reconstruction matrix of 512 × 512.A weight-based (1.5 mL/kg) bolus of nonionic contrast media (Omnipaque, iodine 300 mg/mL; GE Healthcare) was administered intravenously by a power injector at a rate of 3-4 mL/s, followed by a weight-based saline bolus (0.75 mL/kg).The acquisition time was approximately 2 min.No anesthesia or sedating medications were needed.

CT Perfusion Data Analysis
The CT perfusion findings were analyzed to obtain the mean transit time (MTT), cerebral blood volume (CBV), CBF, and time to peak (TTP).The MTT, TTP, and CBF maps were generated using commercial perfusion software (VPCT Neuro, Siemens Healthcare).An arterial input was determined by placing a small circular ROI within the densely enhancing artery.The vein function was selected by placing a circular ROI within a superior sagittal sinus.Hand-drawn ROI placement was performed within the anterior cerebral artery (ACA), middle cerebral artery (MCA), posterior cerebral artery (PCA) circulation territories, and semiovale center.There were two cortical ROIs at different axial levels in the ACA territory in the anterior frontal region, three cortical ROIs at different axial levels in the MCA territory in the perisylvian region, and two cortical ROIs at different axial levels in the PCA territory in the occipital region.The ROI in the semiovale centrum was placed just above the level of the superior border of the lateral ventricle.Absolute average values for each parameter and each region were compared between the preoperative and postoperative sessions.Post-processing stages were performed by two experienced neuroradiologists (5 years of practice).Figure 1 shows the placement of ROI in perfusion maps.

Statistical Analysis
Statistical analysis was performed using the R software (www.r-project.org).The regional values for each hemisphere were analyzed independently.The Wilcoxon signed-rank test was used to compare regional perfusion values of ACA, MCA, PCA circulations, and semiovale centrum before and after surgical intervention.A correlation analysis (Spearman's correlation test) was performed to estimate the relationship between baseline regional perfusion metrics and disease stage, as well as between perfusion metrics and cognitive scores.p < 0.05 (with FDR correction for four comparisons in each metric) was considered statistically significant.The mean values of the variables studied are reported together with ± standard deviation of the mean.

Results
Demographic Data of the Patient, Clinical Information, and Surgical Treatment Ten patients (five boys and five girls aged 5-17 years old; average age: 10.4 ± 3.6 years) participated in this study.Among them, there were 9 patients with bilateral lesions and 1 patient with a unilateral lesion.The moyamoya syndrome was diagnosed in 1 case (thrombophilia was an associated disease).The complete demographic and clinical information on each patient is provided in Table 1.
A total of 13 hemispheres were surgically treated.In all cases, the clinical type of the disease was ischemic (there were ischemic strokes in the clinical history of 4 patients and transient ischemic attacks in 6 patients).The indications for surgery were a clinical history of ischemic events (stroke or transient ischemic attack) and hypoperfusion according to CT perfusion data.Different variants of combined revascularization were used for surgical treatment (Table 1).No intraoperative surgical complications were observed.
One patient (case 5) had partial bypass thrombosis and the development of ischemic stroke, which resulted in impaired speech function on the third day after surgery due to thrombophilia decompensation.After anticoagulant treatment, there was complete recovery of blood flow through the bypass and speech function.Three patients  Partial bypass thrombosis on the third day after surgery due to decompensating of thrombophilia; wound necrosis in the frontal region, complicated with osteomyelitis. 2 Wound necrosis (3 cm in diameter) in the area of sampling the posterior auricular artery.

Brain Perfusion in Moyamoya Angiopathy after Revascularization Surgery
(cases 5, 6, 7) had wound necrosis in the early postoperative period; in one case, an infectious complication developed that required revision surgery.
For 3 years after surgery, none of the patients had clinically significant ischemic events on the side of surgical intervention.There were no anastomosis occlusions or hemorrhagic complications.The 5-to 7-month followup angiographic results of revascularization using the Matsushima classification system were excellent (A) or good (B) in all cases (Table 1).
Three patients had neurological deficits before surgery, which partially regressed 5-7 months after revascularization (Table 1).Cognitive test results were significantly better after surgery in parameters such as the maximum number of words recalled and delayed recall in the 10word test (p < 0.001 and p < 0.05, respectively, FDR corrected), as well as VFT (p < 0.01, FDR corrected).Full information about cognitive testing is shown in Table 2.

Comparison of Regional Brain Perfusion Values before and after Surgery
Brain perfusion parameters improved significantly after surgery in both white matter (semiovale centrum) and gray matter (ACA and MCA circulations).Full results are shown in Table 3 and Figures 2-4.Within the semiovale centrum, the CBV values were lower (p < 0.05, FDR corrected, Fig. 2a), the MTT values were lower (p < 0.05, FDR corrected, Fig. 2b), and the CBF and TTP values did not change significantly after surgery (p > 0.1, not shown).In the ACA and MCA circulation, the CBF values were higher (p < 0.01, FDR corrected, Fig. 3a, 4a), the MTT values were lower (p < 0.01, FDR corrected, Fig. 3b and  4b), the TTP values were lower (p < 0.01, FDR corrected, Fig. 3c, 4c), and the CBV values were not significantly changed after surgery (p > 0.1, not shown).There were no significant differences between the perfusion parameters in the PCA circulation before and after the intervention.

Correlations between Regional Brain Perfusion Values and Suzuki Stage
There were significant positive correlations between the Suzuki stage and the MTT and TTP values within the PCA circulation (R = 0.68, adjusted p < 0.001 and R = 0.86, adjusted p < 0.001, respectively; online suppl.Fig. 1a,  b, for all online suppl.material, see https://doi.org/10.1159/000531719).There were no significant correlations between perfusion metrics within the ACA, MCA circulations, semiovale centrum, and Suzuki stage (p > 0.1, not shown).There were no significant correlations between the percent improvement of the perfusion metric after the operation and the Suzuki stage (p > 0.1, not shown).

Correlations between Brain Perfusion Metrics and Cognitive Functions
Significant negative correlations were found between the CBV values within the semiovale centrum and the maximum number of recalled words and the delayed recall in the Ten-word test (R = −0.75 and R = −0.79,respectively; p < 0.001, FDR corrected, suppl.Fig. 2a, b).Furthermore, there was a significant negative correlation between TTP values in MCA circulation and delayed recall in the Tenword test (R = −0.66,p < 0.05, FDR corrected, suppl.Fig. 2c).There were no significant correlations between the Ten-word test scores and other perfusion metrics (p > 0.1, not shown).There was a significant negative correlation between CBV values within the semiovale centrum and the VFT results (R = −0.81,p < 0.001, FDR corrected, suppl Fig. 3a).Furthermore, there was a significant negative correlation between TTP values in the circulation of MCA and the VFT results (R = −0.62,p < 0.05, FDR corrected, suppl Fig. 3b).There were no significant correlations between VFT results and other perfusion metrics (p > 0.1, not shown).There were no significant correlations between the TMT results and brain perfusion metrics (p > 0.1 after FDR correction, not shown).

Discussion
It has been known for quite some time that perfusion delay is a key pathophysiological event in steno-occlusive conditions, and accurate hemodynamic evaluation is critical to find the best treatment schedule and efficacy assessment in MMA [24].Furthermore, the cerebral hemodynamics of patients with MMA undergo significant changes throughout the natural course of the disease and after surgical treatment.CT perfusion is often utilized in patients with MMA because it is an inexpensive and widely available technique [25].Additionally, this method enables the quantitative assessment of perfusion parameters and is suitable for longitudinal observation and the evaluation of the effect of surgical treatment.Furthermore, CT perfusion metrics can be used as a prognostic and predictive tool for surgical treatment in children with MMA [15].
In this study, we revealed a significant improvement in perfusion metrics within hemispheric white matter (especially in the semiovale center) after bypass surgery, primarily due to a decrease in CBV.On the contrary, gray matter perfusion improved primarily due to an increase in CBF.These findings may help us understand the physiological mechanisms of cerebral hemodynamic changes in patients with MMA after the revascularization Brain Perfusion in Moyamoya Angiopathy after Revascularization Surgery procedure.In general, there is prominent microcirculatory congestion due to compensatory vasodilatation in patients with MMA, resulting in reduced CVR and the appearance of neuroimaging features such as the "brush sign" in SWI and medullary streaks on FLAIR [26,27].In our study, we observed a pronounced decrease in CBV values within the centrum semiovale after surgery, and these changes can be explained by both the improvement in global cerebral perfusion and the "watershed shift" phenomenon.

Brain Perfusion in Moyamoya Angiopathy after Revascularization Surgery
To our knowledge, this is the first study dedicated to differences in brain perfusion between white and gray matter in children with MMA after a revascularization procedure.Han et al. [17] demonstrated an improvement in white matter perfusion parameters in patients with MMA after bypass surgery using whole-brain CT perfusion.Several studies have revealed myelin damage due to hypoperfusion within the hemispheric white matter in patients with MMA [28,29].It is unclear, however, whether these changes are reversible [18,29].
CT perfusion is believed to be a useful diagnostic tool for the evaluation of the bypass surgery effectiveness assessment [30][31][32][33], and our data generally confirm this statement.There were statistically significant improvements in CBF, MTT, and TTP metrics in both ACA and MCA circulations 6-12 months after bypass surgery, and these changes were associated with improvements in cognitive function.
The positive correlation between brain perfusion values and cognitive scores can be explained by improved brain perfusion after bypass surgery.Improvements in memory functioning, as well as VFT results, were associated with improvements in perfusion metrics in the MCA circulation and the semiovale centrum.This can potentially be explained by the general resolution of microcirculatory congestion with subsequent improvement in perfusion in the lateral frontal lobe regions.The observed improvement in TMT results (which probe executive functions) was associated with improvements in perfusion metrics in the ACA circulation; it is well known that the medial frontal lobe is strongly associated with executive functions [34,35].
According to our results, there were also significant positive correlations between the Suzuki stage and the MTT and TTP values within the PCA circulation but not in the ACA or MCA circulations.This is explained by a lesion of P1 segments in patients with late Suzuki stages, as well as blood flow redistribution [17].However, it probably does not have a useful clinical meaning because Suzuki staging represents the hemodynamic reorganization process (internal carotid-external carotid conversion) but not the disease severity [8,36].
This study has several limitations.First, there was a small sample size, which can be easily explained by the low incidence rate of the disease.The second contributing factor is the heterogeneity of the ages of the patients, as well as the fact that both cognitive and perfusion follow-ups were carried out during the CNS maturation process.However, it is known that in the normal pediatric population there are drastic age-dependent CBF dynamics [37,38]; thus, perfusion data interpretation in children with MMA can be complicated.Third, in our study, no acetazolamide challenge was administered.Furthermore, we were unable to compare CT perfusion data with other perfusion methods (such as arterial spin labeling, dynamic susceptibility contrast enhanced magnetic resonance perfusion, or SPECT) in our study.Finally, the retrospective design of our study did not allow for a more detailed assessment of their cognitive status.More studies with larger sample sizes and a prospective design are needed to confirm our results.

Conclusion
In this study, we demonstrated that gray matter and white matter brain perfusion parameters improve differently after bypass surgery in patients with MMA.Therefore, it is important not to underestimate white matter perfusion in this group of patients.

Statement of Ethics
We submitted our study design to the Local Ethics Committee of the Federal Neurosurgical Center, Novosibirsk.The Local Ethics Committee did not identify any ethical concerns as this was an observational study with anonymized data and without any additional therapy or monitoring procedures.According to Russian legislation, the study did not require approval from an ethics committee or written informed consent from patients since it only involved retrospective analysis of prospectively collected anonymized data as part of routine clinical care; each patient was simply informed of his participation in this study and offered the possibility of withdrawing.

Fig. 1 .
Fig. 1.An illustration of the analysis of brain CT perfusion maps: the ROI within the centrum semiovale is taken at the first axial slice above the level of the superior border of the lateral ventricle (marked in red) (a); the ROIs within gray matter are taken at the medial frontal region (ACA circulation, marked in orange), perisylvian region (MCA circulation, marked in red), and occipital region (PCA circulation, marked in green) (b).

Table 1 .
Detailed information about the patients Rankin scale.*Patient with moyamoya syndrome, where thrombophilia was a main disease.

Table 2 .
The full results of cognitive tests