Natural course of partially embolized carotid‐cavernous fistulas

Abstract Background and Purpose To present the first study analyzing the clinical and radiological course of carotid‐cavernous fistulas (CCFs) following incomplete embolization. The study compares magnetic resonance angiography (MRA) to plain angiography (digital subtraction angiography [DSA]) and investigates the long‐term ophthalmological impact of residual fistula. Methods Fistulas classified as partially embolized after the last endovascular treatment were prospectively followed with DSA, MRA, and ophthalmological examination. Both direct and indirect CCFs were included. Results Twenty‐one CCFs were included in the study. Nine (43%) fistulas were direct and 12 (57%) were indirect. A favorable clinical outcome of modified Rankin scale ≤2 was recorded in 19 (90%) patients at the last follow‐up. Postinterventional ophthalmologic examinations in 16 patients revealed no negative effects of residual fistulas; five remaining patients refused to undergo further examination. Spontaneous thrombosis and complete occlusion of the CCF were demonstrated in 90% of patients, with a mean time to occlusion of 5.7 ± 4.7 months. Fourteen (66%) patients completed the full imaging follow‐up (MRA and DSA). In 21% of these cases, discrepancy between the two imaging modalities was observed—MRA failed to detect persistent fistulas identified by DSA. Conclusions The goal of CCF treatment is safe and complete embolization. However, if adequate flow reduction is achieved, both direct and indirect CCFs tend to spontaneously thrombose. Residual flow does not result in ophthalmological deterioration until the fistula is completely closed. MRA may not be sufficiently sensitive to detect residues of fistulas including cortical venous drainage. Therefore, complete CCF closure should be confirmed through DSA.