Prediction of cerebral infarction after bypass surgery in adult moyamoya disease: combing parameters on 4D perfusion CT with clinical related factors

Abstract

Objective: This study aimed to identify predictive factors for cerebral infarction after bypass surgery in adult patients with moyamoya disease (MMD) using quantitative parameters in 4D-CT perfusion software.

Methods: A total of 108 patients who underwent combined revascularization, including superficial temporal artery-middle cerebral artery (STA-MCA) anastomosis and encephalo-duro-myo-synangiosis (EDMS), in our hospital between September 2019 and August 2023 were analyzed retrospectively. Preoperative relative cerebral blood flow (rCBF), relative cerebral blood volume (rCBV), relative mean transit time (rMTT), and relative time to maximum residual function (rTmax) perfusion parameters were obtained using 4D-CT perfusion software. These quantitative parameters, combined with related clinical and angiographic factors, were statistically analyzed using univariate and multivariate regression analyses to determine the significant predictors of cerebral infarction after bypass surgery.

Results: Acute cerebral infarction occurred in 12 patients postoperatively. Univariate analysis showed that a history of previous ischemic events (P = 0.024), higher Suzuki stage (P = 0.006), higher modified Rankin score (mRS) (P = 0.013), rCBV (P = 0.026), rMTT (P = 0.001), and rTmax (P < 0.001) were associated with postoperative cerebral infarction. Further multivariate regression analysis showed that a history of previous ischemic events (OR = 12.830, 95%CI = 1.854-875.672, P = 0.031) and higher rTmax (OR = 16.968, 95%CI = 2.035-141.451, P = 0.009) were independently associated with new postoperative cerebral infarction. The cutoff value for rTmax was 2.025 (AUC = 0.935).

Conclusions: Previous ischemic event history and rTmax greater than 2.025 are independent risk factors for predicting cerebral infarction after combined revascularization with high sensitivity in adult patients with MMD. These patients should be more cautious when deciding on combined revascularization.

Keywords: Adult moyamoya disease; Cerebral infarction; Combined revascularization; RTmax.

Fig. 1

The flowchart showing patient inclusion and exclusion

Fig. 2

The method for obtaining 4D-CT Perfusion parameters preoperatively. We manually marked the semi-oval center and basal ganglia on the surgical side as the region of interest (ROI) and obtained the corresponding perfusion parameters of multiple layers. To standardize the test results, we considered the ipsilateral cerebellar hemisphere-level area as a reference. A shows the ROI delineation method for the perfusion parameter CBF; B shows the ROI delineation method of perfusion parameter CBV; C shows the ROI delineation method of perfusion parameter MTT; D shows the ROI delineation method for perfusion parameter Tmax

Fig. 3

Representative case. A 38-year-old male patient was treated for 2 months for intermittent slurred speech with left limb weakness. There had a history of cerebral infarction 8 weeks before surgery. After admission, the patient was diagnosed with moyamoya disease by DSA examination, Suzuki stage III) by DSA examination (A, B, C). Perfusion parameters at the semi-oval, basal ganglia, and cerebellum levels were obtained using 4D-CTP examination before surgery. After the measurement, aTmax = 9.012 s and rTmax = 3.016 (D). After perfect preoperative preparation, the patient underwent right STA-MCA vascular anastomosis combined with an EDMS. The anastomosis had a good shape during the operation (E). Real-time ICG fluorescence angiography showed that the anastomosis was unobstructed (F). After resuscitation from postoperative anesthesia, the patient was conscious and returned to the intensive care unit without neurological dysfunction. However, the patient suddenly experienced a seizure at night, muscle strength of the left limb decreased, and the patient was in a coma. CTA immediately revealed that the bridge vessels were unobstructed (G), and MRI showed an acute large-area cerebral infarction in the right frontal parietal lobe (H). The patients were then given symptomatic treatment, such as antiplatelet therapy, improved cerebral circulation, dehydration and detumescence, and improved pulmonary ventilation. However, the patient’s condition was not relieved. The MRS score was 5 at discharge and the mRS score was 5 at three-month follow-up

Fig. 4

The ROC curve of rTmax on 4D-CTP. The area under the rTmax curve (AUC) was 0.935 (95%CI = 0.875–0.994, P < 0.001). The threshold was 2.025. At this time, the sensitivity for predicting postoperative cerebral infarction was 91.7% and the specificity was 82.3%

Fig. 5

The correlation between rTmax and Suzuki stage. As shown in the box diagram, the average rTmax values in Suzuki stages I, II, III, IV, and V are 1.221, 1.597 ± 0.540, 1.787 ± 0.686, 2.236 ± 0.865, and 3.094, respectively. With an increase in Suzuki stage, rTmax gradually increased, and there was a positive correlation between them (r = 0.315, two-sided P = 0.001)