Susceptibility-Weighted Angiography versus Arterial Spin Labeling MRI for Transverse-Sigmoid Sinus Dural Arteriovenous Fistula Treatment

Abstract Background Dural arteriovenous fistulas are vascular shunts located within the wall of a venous sinus. Accurate assessment of cortical venous reflux is the strongest predictor of aggressive clinical behavior and guides treatment decisions. Purpose To compare the diagnostic performance of susceptibility-weighted angiography and arterial spin labeling MRI for detecting cortical venous reflux and arteriovenous shunting before and after treatment for transverse-sigmoid sinus dural arteriovenous fistulas. Methods Between January 2012 and December 2024, consecutive patients with intracranial dural arteriovenous fistulas were screened at our institution. Patients with transverse-sigmoid sinus dural arteriovenous fistulas who underwent MRI before and after treatment were retrospectively included. 3T MRI was performed with susceptibility-weighted angiography and pseudo-continuous arterial spin labeling sequences. Two neuroradiologists independently evaluated susceptibility-weighted angiography and arterial spin labeling images. The signal intensities in the Labbe vein, superficial cortical veins, and venous sinus were graded as susceptibility-weighted angiography (1: low; 2: intermediate; 3: high) and arterial spin labeling (1: low; 2: high) images. Grades 2 or higher were defined as venous reflux. Catheter-based digital subtraction angiography served as the reference standard. Results Of the 80 screened patients, 37 were included in the final analysis. Before treatment, susceptibility-weighted angiography demonstrated significantly higher sensitivity than arterial spin labeling for detecting venous reflux in the vein of Labbe (95% [95% CI, 75%–99%] vs 42% [95% CI, 23%–64%]; P = .0011) and superficial cortical veins (72% [95% CI, 54%–85%] vs 28% [95% CI, 15%–46%]; P = .0031), whereas specificity was 100% for both techniques. After treatment, no significant differences in sensitivity or specificity were observed between the two techniques. Conclusion Susceptibility-weighted angiography may be more sensitive than arterial spin labeling for detecting cortical venous reflux in transverse-sigmoid sinus dural arteriovenous fistulas before treatment, while arterial spin labeling and susceptibility-weighted angiography may be comparable for residual shunting and cortical venous reflux after treatment. These modalities are complementary and may together improve noninvasive evaluation and follow-up of dural arteriovenous fistulas.