Introduction/purpose Indirect Carotid-Cavernous Fistulas (CCF) are arterio-venous shunts between branches of the ICA, ECA or both and the cavernous sinus. Indirect CCFs are treated with coil embolization when they present with orbital/visual symptoms or if there is cortical venous reflux. The target for endovascular treatment in indirect CCF is the occlusion of the venous side of the fistula. Due to intricate anatomy of the cavernous sinus, the fistula site and foot of the draining venous structure are difficult to localize with confidence. This makes treatment a complex and long procedure requiring a large number of coils. Most of the time the treatment is done without actually localizing the fistula site and by packing the whole cavernous sinus with coils, sometimes resulting in additional complications, long procedure times and a large number of coils used per procedure.

The purpose of this study was to examine the role of DynaCT in pre-treatment localization of the fistula site for indirect CCF, and to examine if identification of fistula site would improve treatment by shortening the procedure time, requiring fewer coils or by reducing complication rate.

Methods Patients with endovascular treatment of indirect CCF between 2005–2015 were retrospectively identified and reviewed from our institutional database. The patients were assessed to see if DynaCT was used in the planning of treatment to identify the site of fistula. All patients were divided into two groups- those who had DynaCT used for treatment planning and those who did not. These two groups were further compared in a retrospective cohort study.

Results A total of 8 patients with 9 fistula sites were included, with a mean age of 52.3 years (range 40–67 years). The majority of the fistulas were Barrow Type D (75%), and 62.5% of cases also had cortical venous reflux. DynaCT was used in 5/8 (62.5%) patients. Transfemoral venous access to the fistula itself was possible in 7/8 patients (87.5%). One patient required direct surgical access via ophthalmic vein cut down. 7/8 patients reported resolution of symptoms at clinical follow-up (87.5%).

Mean total coil length was significantly shorter for the group who had DynaCT than for the group who did not (156.4 cm, vs. 190 cm, p = 0.035). Mean procedural time was 209 minutes for the DynaCT group vs. 280 minutes for the Non-DynaCT group (p = 0.12). A transient neurological complication was seen in only one of the patients.

Conclusion The use of DynaCT in pre-procedural planning allows identification of fistula site for indirect CCF and facilitates selective coil embolization of the foot of the vein. This results in the use of significantly shorter total coil length.

Disclosures A. Botsford: None. J. Shiva Shankar: None.