Abstract No. 129 Endovascular Management of Visceral Vein Obstruction: Technical Considerations and Factors Associated with Failure

To report the outcomes of recanalization of long segment, visceral vein occlusion (VVO) in patients with or without cirrhosis. All attempted visceral venous recanalizations (VVR) were retrospectively reviewed at a single academic center. Procedural data regarding technique, including balloon angioplasty, stent placement, and transjugular intrahepatic portosystemic shunt (TIPS) creation, was collected. Patency was recorded during follow up and compared in groups who did and did not have a TIPS placed. 40 patients who underwent 45 attempted recanalization procedures of which 36 (36/45,80%) were successful and 9 (9/45,20%) failed were included. A total of 78 anatomically separate veins were occluded, with main portal veins being the most common site of occlusion (32/78,41%). The median length of occlusion across affected vessels was 6.5 cm (IQR=5.51 cm). The median time of occlusion from diagnosis to recanalization was 5 months (IQR=21.5 months). Follow-up imaging was performed for a median duration of 13.5 months (IQR=12.5 months). The median duration of primary patency was 5.2 months (IQR=24.14 months). Univariate analysis demonstrated that occlusion length (OR:-0.26(95%CI:-0.40,-0.12),p<0.001) and stent placement not including TIPS) (OR:1.6(95%CI:0.17,3.0) p= 0.029) were significantly associated with patency. Thirty-one patients (31/40,77.5%) had a TIPS placed before, after, or at the time of recanalization. Kaplan-Meier curves showed that patency (p=0.18) and survival (p=0.74) of patients who had VVR with/without TIPS were not statistically significantly different. All recanalized patients had resolution of symptoms or underwent successful abdominal surgery. VVR is technically achievable in long-segment, long-term occlusions, and should be considered for patients who are symptomatic.