TRICKS Magnetic Resonance Angiography at 3-Tesla for Assessing Whole Lower Extremity Vascular Tree in Patients with High-Grade Critical Limb Ischemia: DSA and TASC II Guidelines Correlations

Abstract

The entire vascular tree of 58 lower extremities with high-grade critical limb ischemia (CLI) was assessed with three-station time resolved imaging of contrast kinetics (TRICKS) magnetic resonance angiography (T-MRA) and correlated with digital subtraction angiography (DSA) examinations and Trans-Atlantic Inter-Society Consensus II (TASC II) guidelines. Kappa (κ) statistics were utilized to evaluate the agreement of stenosis scores (5-point scale; 0 normal to 4 occlusion) based on T-MRA and DSA. With DSA as the standard, significant stenosis instances (stenosis score ≥2) among vascular segments were compared. The κ-statistics of image quality (4-point scale; 1 nondiagnostic to 4 excellent) of T-MRA and TASC II classification assessed by a radiologist and a vascular surgeon were also evaluated. Among 870 vascular segments, excellent agreement was observed between T-MRA and DSA (mean κ = 0.883) in revealing stenosis (mean stenosis score, 2.1 ± 1.3 versus 2.0 ± 1.3). T-MRA harbored overall high sensitivity (99.5%), specificity (93.6%), positive predictive value (95.4%), negative predictive value (99.6%), and accuracy (97.7%) in depicting significant stenosis. Excellent interobserver agreement (mean κ = 0.818) of superb image quality (mean score = 3.5–3.6) of T-MRA and outstanding agreement of TASC II classification of aortoiliac and femoral-popliteal lesions (κ = 0.912–0.917) between two raters further verified the clinical feasibility of T-MRA for treatment planning.