Perioperative ASL-derived CBF changes are associated with complications after Moyamoya bypass surgery

Postoperative hyperperfusion and insufficient perfusion augmentation are important contributors to morbidity after revascularization surgery for Moyamoya disease (MMD). Arterial spin labeling (ASL) enables noninvasive assessment of cerebral blood flow (CBF), but the relationship between perioperative ASL-derived CBF changes and specific postoperative complications after direct bypass remains incompletely defined. This study evaluated whether perioperative ASL-derived CBF changes were associated with postoperative complications after direct bypass surgery in patients with MMD.

We retrospectively reviewed 146 patients with MMD who underwent first-time direct bypass surgery between June 2020 and October 2024. Preoperative ASL and a standardized postoperative ASL examination on postoperative day 5 were performed, and CBF values were measured in the recipient vascular territory. Subsequent intracranial hemorrhage, cerebral infarction, and postoperative seizures occurring within 90 days after surgery were analyzed. Firth penalized logistic regression was used for the hemorrhage outcome, conventional logistic regression was used for cerebral infarction, and ROC analyses included bootstrap assessment of cutoff uncertainty and 0.632 + bootstrap correction of AUC optimism.

Postoperative intracranial hemorrhage occurred in 8 patients (5.5%), cerebral infarction in 14 patients (9.6%), and postoperative seizures in 9 patients (6.2%); all events occurred after the day-5 ASL examination and within 90 days after surgery. ASL-derived CBF change was associated with postoperative intracranial hemorrhage in Firth penalized logistic regression (OR, 1.219; 95% profile-likelihood CI, 1.102-1.580; p < 0.001). The exploratory cutoff was >43.48 mL/100 g/min, with an apparent AUC of 0.9891 and a 0.632 + corrected AUC of 0.9890. ASL-derived CBF change was inversely associated with postoperative cerebral infarction (OR, 0.885; 95% CI, 0.839-0.933; p < 0.001). The exploratory cutoff was ≤4.48 mL/100 g/min, with an apparent AUC of 0.9150 and a 0.632 + corrected AUC of 0.9150. No significant association was observed between ASL-derived CBF change and postoperative seizures.

Perioperative ASL-derived CBF change was associated with hemorrhagic and ischemic complications after direct bypass surgery for MMD. Marked postoperative CBF elevation was observed in patients with intracranial hemorrhage, whereas insufficient perfusion augmentation was associated with cerebral infarction. The proposed thresholds are exploratory, cohort-specific estimates and require external validation before clinical application.