Background: The presence of a cortical venous reflux (CVR) in a carotid-cavernous fistula (CCF) is well described and is considered to be a criterion for urgent treatment. This reflux is often associated with direct/traumatic CCF and the high-flow of the fistula alone explains the reflux. For indirect CCF, the pathophysiology of a CVR is unclear.
Methods: All patients treated endovascularly for an indirect CCF with a cortical venous reflux between 2003 and 2015 were included. We retrospectively analysed data focusing on whether venous outflows of the cavernous sinus would opacify or not with the local injection of contrast, in order to locate those that could explain the venous reflux.
Results: Twenty consecutive patients (male/female ratio, 2/3) were included in this series with a mean age of 63 years. All patients presented ocular signs and no patients showed any neurological sign secondary to the CVR. We distinguished four patterns of CVR: in the superficial middle cerebral vein (75%), in the uncal vein (15%), in the superior petrosal vein (5%) and in the inferior petrosal vein (5%). Seventy percent of the cases presented a lack of opacification in more than three venous outflows of the CS involved. Each patient received an endovascular therapy by venous approach with a success rate of 76.9% per embolisation session. Two patients (10%) presented a permanent ocular paresis and two others a transient deficit.
Conclusions: CVR is directly correlated with the thrombosis of multiple venous outflows of the CS. The "non-opacification" of at least three of the CS venous outflows is necessary for the development of CVR. Such thrombosis may be explained by the combination of haemodynamic and inflammatory changes of the venous wall.
Keywords: Carotid-cavernous fistulas; Cerebral angiography; Cortical venous reflux; Endovascular therapy.