Purpose This project sought to test the utility of post-delivery human placenta (HP) as a vascular model for liquid embolic agent (LEA) simulation, along with adjunctive techniques.

Materials and methods Twelve LEA injections were performed under fluoroscopy in HP with two reflux control methods: dual lumen ‘mini’ balloon-catheter (n=9); and injection after proximal nBCA plug formation through a second microcatheter (‘pressure cooker’) (n=3). Measured outcomes included liquid embolic agent (LEA) advancement and reflux. Reflux was categorized into three grades: grade 0=no reflux; grade 1=occlusion of side branches without reflux beyond the balloon or plug; and grade 2=reflux beyond the balloon or plug.

Results Simulation success was greater when a balloon was used rather than with a nBCA plug (89% vs 33%, P=0.054). In eight successful balloon-assisted injections, the reflux grades were: 50% grade 0; 12.5% grade 1; and 37.5% grade 2. The one successful nBCA plug injection had grade 2 reflux. All grade 2 balloon injections occurred when the balloon was positioned across a vessel bifurcation.

Conclusions HP provides excellent simulation for liquid embolic agents with a dual lumen balloon catheter.