Network meta-analysis of endovascular treatments for lower extremity arterial disease stratified by lesion location and severity.

Abstract

Although numerous endovascular technologies are available for lower extremity arterial disease (LEAD), a lack of direct, comparative evidence complicates optimal device selection. We conducted a network meta-analysis of randomized controlled trials (RCTs) to establish an evidence-based framework for device selection across distinct clinical-anatomical subgroups. Following the PRISMA guidelines (PROSPERO: CRD42024629620), major databases were searched for RCTs comparing nine endovascular treatments for LEAD. The analysis was stratified into four subgroups based on lesion location and clinical severity (intermittent claudication [IC] vs critical limb ischemia [CLI]). Primary outcomes were primary patency, target lesion revascularization, major amputation, and all-cause mortality. We performed a random effects network meta-analysis with sensitivity analyses to assess for multiple comparisons. A total of 82 RCTs involving 12,796 patients were included. For femoropopliteal lesions in IC, a time-dependent hierarchy was observed: atherectomy combined with drug-coated balloon (DCB) excelled in short-term (6-month) patency, while covered stents (CS) and drug-eluting stents (DES) were superior at 24 and 36 months, respectively. For iliac lesions in IC, CS showed higher patency rates than bare metal stent (BMS) and balloon angioplasty (BA) at 12 months. In infrapopliteal CLI, drug-eluting technologies (DES and DCB) were superior for both patency and target lesion revascularization at 12 months, with DES associated with a lower 12-month major amputation rate compared with DCB (odds ratio, 0.41; 95% confidence interval [CI], 0.18-0.94). For femoropopliteal CLI, time-to-event analysis revealed that BA resulted in lower primary patency compared with both BMS (hazard ratio, 0.44; 95% CI, 0.21-0.92) and DCB (hazard ratio, 0.69; 95% CI, 0.49-0.98) over a 12- to 24-month period. Across all comparisons, no significant differences in all-cause mortality were found. A sensitivity analysis using 99% CIs confirmed the robustness of most findings, but the superiority of CS in iliac lesions and the amputation benefit of DES over DCB became nonsignificant, warranting caution. Endovascular device selection for LEAD should be tailored to the clinical-anatomical context. For IC, CS trends toward being the preferred option for iliac lesions, whereas a dynamic hierarchy of devices exists for femoropopliteal disease, balancing long-term patency against reintervention risk. For CLI, drug-eluting technologies are paramount for infrapopliteal disease, with DES showing a potential but not robust 12-month advantage in limb salvage. For femoropopliteal CLI, evidence suggests that both stenting (BMS) and drug-eluting technologies (DCB) offer superior patency over 12 to 24 months compared with BA alone.