Background: Lower extremity revascularization (LER) is frequently used to treat symptomatic peripheral artery disease (PAD). Factors influencing the probability of LER in symptomatic PAD and subsequent outcomes are poorly defined.
Methods: We describe characteristics of patients in the EUCLID trial who underwent post-randomization LER and the types of procedures. Patients were grouped as having a post-randomization LER or not. All variables were assessed for significance in univariate and multivariate cox models with LER as the outcome. Rates were computed as number of events per 100 patient-years of follow-up.
Results: A total of 1,738 patients (14.3%) in EUCLID had post-randomization LER. Most had endovascular LER (n=1297; 74.6%) in the iliac (30.0%) and superficial femoral arteries (31.5%). In those with surgical procedures (n=440 [25.3%]), the most common were femoropopliteal bypass surgery (40.7%) and femoral artery endarterectomy (27.5%). Those undergoing LER were more likely to smoke, have diabetes, hypertension, prior history of LER, co-existent coronary and/or carotid disease, and had significant (p<0.001) regional variation with the highest number coming from North America. Independent factors predicting LER included prior antiplatelet use, limb symptoms (HR 1.32; 95% CI 1.16-1.49), current or former tobacco use (HR 1.25; 1.08-1.44), diabetes (HR 1.23; 1.11-1.36), type of prior LER (p<0.0001), and region (p<0.0001). Event rates (n/100 pt-yrs) comparing type of procedure (surgical LER vs. endovascular LER) post-randomization were 11.3 vs 20.5 for a subsequent LER after the index procedure; 4.78 vs. 4.56 for the primary efficacy endpoint of CV death, MI, stroke; 2.03 vs. 1.51 for CV death; 3.08 vs. 1.84 for ALI requiring hospitalization; 4.27 vs. 1.39 for major amputation; and 0.93 vs. 0.40 for major bleeding.
Conclusion: Risk modeling demonstrated presence of limb symptoms, diabetes, current or former tobacco use, type of previous LER, and region are significantly associated with LER. Major outcome events (CV death, ALI requiring hospitalization including major amputation) was observed more frequently in surgical LER, whereas repeat LER procedure were more frequent in patients after endovascular LER.