Direct carotid cavernous sinus fistulae: vessel reconstruction using flow-diverting implants.

Purpose: Retrospective evaluation of our experience with the use of flow diverters (FD) for the endovascular treatment of direct carotid-cavernous sinus fistulae (diCCF).
Methods: Between 2011 and 2015, 14 consecutive patients with 14 diCCF were treated with FD alone or in combination with other implants in a single institution.
Results: A total of 21 sessions were performed in 14 patients. FD placement was technically successful in all cases without an adverse event. Patients were treated with FD alone (n = 5), FD and covered stents (n = 2), FD and coils (n = 7). A total of 59 FD (24 Pipeline Embolization Device, Medtronic; 35 p64 Flow Modulation Device, phenox), 291 coils, and 3 stent grafts were used. Three of 14 diCCF were completely occluded after the 1 ^st  session, a minor residual shunt was found in 7/14, and in the remaining 4/14 patients, the shunt volume was reduced significantly. The mean follow-up period encompassed 20 months. Additional treatment included transvenous coil occlusion (n = 3) and/or further FD deployment (n = 5). An asymptomatic internal carotid artery (ICA) occlusion was encountered in 2 patients, related to an interruption of antiaggregation. At the last follow-up, 10/14 patients were free from ocular symptoms (71 %), 2 had residual exophthalmos, and no patient had clinical deterioration.
Conclusion: The usage of FD for the treatment of diCCF is straightforward. Injury of the cranial nerves can be avoided. In most cases, ocular symptoms improve. Several FD layers and/or an adjunctive venous coil occlusion are required. Complete occlusion of a diCCF may take weeks or months and long-term antiaggregation is required. In the future, a flexible stent graft might be a better solution.