Purpose: To evaluate the effectiveness of proximal embolization of the gastroduodenal artery (GDA) using the Amplatzer Vascular Plug 4 (AVP 4) compared with pushable coils to avoid hepaticoenteric collaterals of the GDA stump, which may serve as pathways for nontarget embolization.
Materials and methods: One hundred thirty-four patients scheduled for 90-yttrium radioembolization (Y-90 RE) using either plugs (n = 67) or standard coils (n = 67) for GDA occlusion were retrospectively analyzed. Parameters recorded were length of the perfused GDA stump, distance device to the GDA origin, perfused proximal side branches after embolization, and durability of vessel occlusion at Y-90 RE.
Results: Length of the residually perfused GDA stump was 3.89 ± 2.86 mm for the AVP 4, which was significantly shorter compared with 5.78 ± 3.85 mm for coils (p = 0.005). Distance of the plug to the GDA origin was 1.41 ± 2.60 mm, which was also significantly shorter than 4.73 ± 3.44 mm for coils (p < 0.001). This resulted in significantly fewer patients with residually perfused side branches in the AVP 4 group (n = 2; 3.0%) compared with the coil group (n = 18; 26.9%; p < 0.001). At Y-90 RE, no GDA reperfusion was found after plug embolization compared with 2 cases after coil embolization (3.0%; p = 0.156). Only one patient had a radiation-induced duodenal ulcer after coil embolization, whereas no Y-90-related toxicity was identified after plug embolization.
Conclusion: Use of the AVP 4 for embolization of the GDA allowed an optimal proximal and more effective target vessel occlusion compared with coil embolization, which can avoid complications caused by extrahepatic gastrointestinal deposition of Y-90 microspheres by way of residually perfused proximal side branches.