OBJECTIVE: Abducens nerve palsy (ANP) after transvenous embolization (TVE) for cavernous sinus (CS) dural arteriovenous fistulas (DAVFs) sometimes persists. The aim of this study was to assess the correlation between the coil mass and the long-term outcome of ANP after TVE. METHODS: Between January 2008 and July 2016, 33 patients with a CS DAVF underwent TVE at our institute. On the basis of the relationship to the internal carotid artery (ICA), we divided the lateral part of the CS into the following 3 portions: anterolateral, anterolateral to the anterior bend of the ICA; middle-lateral (ML), lateral to the horizontal segment of the ICA; and posterolateral, posterolateral to the posterior bend of the ICA. RESULTS: ANP persisted in 4 patients. The number of coils (35.3 vs. 21.9 coils; P = 0.04), coil volume (198.4 vs. 103.6 mm3; P = 0.03), and coil volume in the ML (54.9 vs. 20.3 mm3; P = 0.01) were significantly greater in the ANP group than in the non-ANP group. In the logistic regression analysis, only the ML coil volume was significantly associated with the persistence of ANP (P = 0.04). Based on the receiver operating characteristic curve, the optimal cutoff value of the ML coil volume was 27.9 mm3 (sensitivity, 100%; specificity, 72.4%). CONCLUSIONS: Overpacking in the ML of the CS should be avoided to prevent persistent ANP. The cutoff value of the ML coil volume may provide a good guide for the practical use of TVE for CS DAVFs.
OBJECTIVE:Abducens nerve palsy (ANP) after transvenous embolization (TVE) for cavernous sinus (CS) dural arteriovenous fistulas (DAVFs) sometimes persists. The aim of this study was to assess the correlation between the coil mass and the long-term outcome of ANP after TVE. METHODS: Between January 2008 and July 2016, 33 patients with a CS DAVF underwent TVE at our institute. On the basis of the relationship to the internal carotid artery (ICA), we divided the lateral part of the CS into the following 3 portions: anterolateral, anterolateral to the anterior bend of the ICA; middle-lateral (ML), lateral to the horizontal segment of the ICA; and posterolateral, posterolateral to the posterior bend of the ICA. RESULTS: ANP persisted in 4 patients. The number of coils (35.3 vs. 21.9 coils; P = 0.04), coil volume (198.4 vs. 103.6 mm3; P = 0.03), and coil volume in the ML (54.9 vs. 20.3 mm3; P = 0.01) were significantly greater in the ANP group than in the non-ANP group. In the logistic regression analysis, only the ML coil volume was significantly associated with the persistence of ANP (P = 0.04). Based on the receiver operating characteristic curve, the optimal cutoff value of the ML coil volume was 27.9 mm3 (sensitivity, 100%; specificity, 72.4%). CONCLUSIONS: Overpacking in the ML of the CS should be avoided to prevent persistent ANP. The cutoff value of the ML coil volume may provide a good guide for the practical use of TVE for CS DAVFs.