This study was approved by the institutional medical ethics committee, and informed consent was waived. Data of 245 consecutive patients with MMD who underwent STA-MCA anastomosis or EDAS between April 2017 and August 2022 were retrospectively analyzed. The diagnostic criteria for MMD were based on the guidelines published in 2022 [2]. The inclusion criteria of patients were as follows: (1) age ≥ 18 years, (2) complete pre- and postoperative clinical data, (3) complete pre- and postoperative CTP and DSA examinations, and (4) follow-up at least 6 months after surgery. Patients with MMD who underwent combined surgical procedures or bilateral hemispheric surgery were excluded. Eighty-four patients were enrolled (Fig. 1). During the follow-up period, none of the patients in either group has experienced a new ischemic or hemorrhagic event.

Flowchart of the study. MMD, Moyamoya disease; STA-MCA, superficial temporal artery-middle cerebral artery; DSA, digital subtraction angiography; EDAS, encephaloduroarteriosynangiosis; CTP, computed tomography perfusion

Surgeries were performed by an experienced neurosurgical expert (F.J. with 25 years of experience in neurosurgery). STA-MCA anastomosis refers to the end-to-side anastomosis of the dissected frontal branch or apical branch of the superficial temporal artery with the M4 MCA segment [9]. EDAS refers to the dissected frontal branch or apical branch adhesions in the superficial temporal artery to the arachnoid mater, with its edge sutured to the dura mater [22].

The principles of the surgical strategies at our institution were as follows. First, the selection of surgical hemispheres is mainly based on clinical manifestation, and symptomatic hemisphere was the preferred side for revascularization surgery. Second, for patients without an obvious symptomatic hemisphere, CT perfusion shows that the hemisphere with more severe hypoperfusion injury is the preferred treatment hemisphere. Third, STA-MCA was the favored surgical modality for most patients, whether in ischemic or hemorrhagic patients. An EDAS was performed only when the donor or recipient artery was too small or fragile to perform artery anastomosis.

All CTPs were performed on a 128-slice CT scanner (SOMATOM Definition Flash, Siemens Healthineers) using a MEDRAD high-pressure syringe. After a non-contrast brain scanning, 50 mL of iopromide (Ultravist 300 or 370; Bayer Schering) was injected for multiphase dynamic scanning. Reconstructed CTP images were transferred to the multimodality postprocessing station (Syngo MMWP, VE 40C, Siemens Healthcare) using the VPCT perfusion software to obtain various CTP parameters: cerebral blood flow (CBF), cerebral blood volume (CBV), time to peak (TTP), and mean transit time (MTT). These images were independently analyzed by 2 neuroradiologists (D.L. and Y.C. with 18 and 22 years of experience in neuroradiology, respectively) who were blinded to the clinical data, type of surgery, and DSA results, and the CTP interpretation training program was performed for both to ensure the consistency of interpreting the standard. The first artery to reach peak enhancement on the time-attenuation curve was selected as the arterial input function. The placement of the regions of interest (ROIs) should avoid the large blood vessels, artifacts, hemorrhage, and infarct area. In addition to the operative area, ROIs of the frontal lobes and basal ganglia, representing the blood supply area of the anterior cerebral artery and deep brain tissue, were also conducted, respectively [23,24,25] (Fig. 2). In our study, the relative CTP values were defined as absolute CTP values in the surgical side divided by absolute CTP values in the pons [26]. The change in postoperative perfusion was defined as differences between relative parameters pre- and postoperatively.

Example of regions of interest (ROIs) of the frontal lobe, operative area, and basal ganglia, drawn in the reference CT image (a) and time-to-peak map (b)

DSA images were acquired using transfemoral cerebral angiography, such as superselective angiography of bilateral external carotid arteries. The Matsushima grading system [27] was used to classify the extent of collateralization: grade 0 refers to no evident collateralization, and grades 1, 2, and 3 are defined as < 1/3, between 1/3 and 2/3, and > 2/3 of the MCA territory perfused by collateralization, respectively.

Clinical outcomes were evaluated by two neurosurgical experts (G.S. and H.Z. with 22 and 20 years of experience in neurosurgery, respectively), both blinded to the result of imaging and details of surgery, through the patient files. Any inconsistency was resolved by senior supervisors’ consultation and thorough discussion on roundtable conference with the Academic Committee. Refer to previous reports [28,29,30], the change in clinical symptoms was classified into four levels: excellent, preoperative symptoms (such as transient ischemic attacks or seizures) had totally disappeared without fixed neurological deficits; good, symptoms had totally disappeared, but neurological deficits remained; fair, symptoms persisted, albeit less frequently; and poor, symptoms remained unchanged or worsened. “Excellent” and “good” were considered as favorable remission of clinical symptoms. Moreover, neurological status was evaluated using the Modified Rankin Scale (mRS) score either.

Intra- and inter-observer correlations for the CBF values were analyzed by estimating the ICCs (the degree of agreement is interpreted as poor for ICC < 0.5, moderate for ICC 0.5–0.75, good for ICC 0.75–0.9, and excellent for ICC > 0.9) [31]. The Statistical Package for the Social Sciences 25.0 statistical software was used for data analysis. Normally distributed measurement data are expressed as mean ± standard deviation (\(\overline\) ± SD) and compared using the two-sample t-test and paired t-test. Non-normally distributed measurement data are expressed as the median (interquartile range [IQR]) and compared using Wilcoxon’s rank-sum test. Count data are expressed as number of patients (percentage) [n (%)] and compared using the chi-square (χ2) test or Fisher’s exact probability method. Grade data were compared using Wilcoxon’s rank-sum test. Spearman rank correlation analysis was used to evaluate the correlation between hemodynamic improvement and clinical symptom remission. p < 0.05 was considered statistically significant.