Caroticocavernous Fistulae: A Real World Comparison of Clinicoradiographic Features and Treatment Approaches
Justin Vranic1, Robert Regenhardt1, Amine Mohamed Marc Awad2, James Rabinov1
1Massachusetts General Hospital, 2Neurology, Massachusetts General Hospital
Objective:
This study aimed to describe the clinical outcomes in a cohort of caroticocavernous fistulae (CCF) patients and identify clinical and angiographic features differentiating patients more likely to benefit from endovascular intervention.
Background:
 CCFs are a rare subtype of intracranial arteriovenous fistulae with variable clinical and angiographic presentations that make treatment selection challenging.
Design/Methods:

A single-center, retrospective analysis was performed to identify all patients with angiographically-confirmed CCF between 2000 and 2022. Pertinent data, including clinical symptoms, angiographic findings, treatment strategies, recurrence rates, and complications, were collected.

Results:

A total of 84 patients were included, of whom 67 (80%) underwent endovascular intervention and 17 (20%) were conservatively managed. Primary endovascular techniques were transvenous coil embolization (78%), feeder artery embolization (16%), and ICA flow diversion (8%). High-risk clinical symptoms, such as reduced visual acuity (53% of intervention vs 6% of conservative, p=<0.0001), and angiographic features, such as cortical (39% vs 0%, p=0.002) and ophthalmic venous reflux (91% vs 69%, p=0.034), were more common in the intervention group. All direct (Barrow Type A) CCFs underwent endovascular intervention (32% vs 0%, p=0.005), while indirect (Barrow types B-D) CCFs were common in the conservatively managed group (68% of intervention vs 100% of conservative, p=0.005). 31% of treated CCFs required retreatment, which primarily occurred with Barrow type D CCFs and following transvenous coil embolization as the initial treatment method. Procedure-related complications occurred in 10% of cases and consisted of cranial nerve palsies (n=5), asymptomatic dissection (n=1), and asymptomatic distal thromboembolic event (n=1).

Conclusions:
High-risk symptoms and angiographic features favor endovascular intervention. Various embolization techniques, including transvenous coil embolization, proved safe and effective; complications were rare and mostly transient. In select CCF patients without high-risk clinical or angiographic features, conservative observation was a safe and effective alternative to endovascular embolization. Longitudinal angiographic surveillance is essential for monitoring fistula persistence or recurrence.
10.1212/WNL.0000000000204622