Intravascular ultrasound-based decision tree model for the optimal endovascular treatment strategy selection of femoropopliteal artery disease—results from the ONION Study-

Abstract

Abstract Background The role of catheter-based imaging in peripheral interventions for lower extremity artery disease (LEAD) has increased with percutaneous interventions. To clarify the relation between intravascular ultrasound (IVUS) information and procedure selection strategy for endovascular treatment therapy (EVT) of the femoropopliteal artery in the real-world clinical settings wherein new endovascular technologies (NETs), including drug-coated balloon (DCB), drug-eluting stent (DES), and covered stent-graft (CS). Our retrospective multicenter analysis examined symptomatic 970 patients treated by EVT for de novo femoropopliteal lesions with IVUS guidance. The decision tree analysis was performed retrospectively to determine the association of IVUS and angiography parameters with the strategy selection of endovascular procedures. We divided the study population according to the developed tree, and identified the most popular strategy selection in each subgroup. We finally examined whether the restenosis risk would be different among respective subgroups of the tree. Results During the study periods, plain old balloon angioplasty, DCB, and bare nitinol stent were most frequently selected (25.3%, 23.9%, and 23.8%, respectively). The drug-eluting stent (DES), covered stent (CS), and spot stent strategies were used in 7.3%, 11.5%, and 8.1%. NETs had the lowest restenosis risk in the overall population. The decision tree had a depth of six branches and divided the patients into 11 subgroups by IVUS and angiography parameters. The restenosis rate was similarly low among these 11 subgroups when the most popular NET in each subgroup was selected (P = 0.94). Conclusions The use of IVUS data along with angiography data would standardize the selection of endovascular procedures and can improve patency outcomes if NETs are used properly.