Intraoperative Cortical Indocyanine Green Extravasation as a Predictor of Cerebral Hyperperfusion following Direct Revascularization for Moyamoya Disease- Impact of Prolonged Observations of ICG Videoangiography.

Abstract

Introduction: Local vasogenic edema following direct revascularization for Moyamoya disease (MMD) is considered to result from an intrinsic vulnerability of blood-brain-barrier (BBB) and is consequently associated with transient focal cerebral hyperperfusion (CHP). However, intraoperative identification of the local vasogenic edema remains challenging. To address this, we implemented a prolonged observation of indocyanine green videoangiography (ICG-VAG) as an extension of routine clinical practice. This approach aimed to investigate intraoperative cortical ICG extravasation following direct revascularization, as an indicator of BBB dysfunction in patients with MMD, providing real-time intraoperative evidence of compromised vascular integrity.

Methods: This prospective observational study included 50 consecutive combined direct/indirect revascularization surgeries performed for MMD at our institution between December 2022 and February 2025. After confirming the patency of the direct anastomosis using ICG-VAG in the early phase, we conducted an additional observation of ICG-VAG in the late phase, approximately five minutes after the initial assessment, to evaluate cortical changes around the anastomotic site. We analyzed the correlation between intraoperative cortical ICG extravasation observed in the late phase of ICG-VAG and postoperative complications, including CHP.

Results: ICG-VAG confirmed patent direct anastomoses in all 50 revascularizations, and postoperative CHP occurred in 16 surgeries (32%) between postoperative days one and seven. Among these cases, intraoperative cortical ICG extravasation was detected in the late phase of ICG-VAG in nine of 50 surgeries (18%). This extravasation presented as focal or patchy leakage of ICG dye near the anastomotic site and/or flow-augmented cortical areas without evident cortical contusion or subarachnoid hemorrhage under the light-field surgical microscope. There were no significant differences in preoperative baseline clinical characteristics between patients with and without cortical ICG extravasation. However, intraoperative cortical ICG extravasation was significantly correlated with postoperative CHP (odds ratio: 12; 95% confidence interval: 2.5-94; P=0.0044) and local vasogenic edema on magnetic resonance imaging (odds ratio: 20; 95% confidence interval: 2.2-444; P=0.015).

Conclusion: Intraoperative cortical ICG extravasation, observed in the late phase of ICG-VAG, may serve as a direct indicator of the intrinsic vulnerability of BBB in patients with MMD. Prolonged ICG-VAG observation could be a simple and effective intraoperative tool to predict postoperative CHP and local vasogenic edema in patients undergoing direct revascularization for MMD, thereby enabling intensive postoperative monitoring for high-risk cases.