Diagnostic Accuracy of Noncontrast MR Angiography Protocols at 3T for the Detection and Characterization of Lower Extremity Peripheral Arterial Disease

Abstract

PurposeTo compare the diagnostic accuracy of established non–gadolinium (Gd)-enhanced magnetic resonance (MR) angiography protocols with Gd-enhanced MR angiography at 3T for evaluating lower extremity peripheral arterial disease (PAD). Materials and MethodsFrom February 2014 to 2015, 20 patients with PAD and intermittent claudication (16 men; age range, 51–76 y; Fontaine stage II) underwent 3-station (abdominopelvic, thigh, and calf) non-Gd MR angiography and bolus-chase Gd MR angiography protocols performed at 3T (Siemens Tim Trio), including quiescent-interval single-shot (QISS) MR angiography for all 3 stations and a combination of quadruple inversion recovery (QIR) MR angiography for the abdominopelvic station and electrocardiogram-gated fast spin echo (ECG-FSE) MR angiography for the extremities. Two radiologists independently evaluated vessel segments for vascular stenosis, diagnosis confidence, graft presence, and Trans-Atlantic Inter-Society Consensus (TASC) II classification for each station. Diagnostic accuracies and κ agreement were assessed. ResultsOf 573 vascular segments imaged, 16.9% (97/573, 19/20 patients) demonstrated hemodynamically significant abnormalities. Reader confidence was sufficient for diagnosis in 98% of segments with Gd MR angiography, 93% with QIR/ECG-FSE, and 95% with QISS. Overall reader confidence was higher with QISS than QIR/ECG-FSE within all 3 stations combined (P < .05). With low-confidence segments treated as misdiagnosis, sensitivity, specificity, positive predictive value, negative predictive value, accuracy, and κ agreement for all 3 stations combined were 81.4/87.2/57.0/95.8/86.2%/0.578 for QIR/ECG-FSE and 75.0/90.6/61.6/94.7/88.0%/0.597 for QISS. Using TASC II criteria to assess severity, QISS and QIR/ECG-FSE had no statistical difference in agreement with Gd MR angiography. ConclusionsQISS and QIR/ECG-FSE MR angiography protocols demonstrate comparable diagnostic accuracies with high specificity. Either protocol provides an alternative to Gd MR angiography at 3T for patients with PAD.