ECG-triggered non-enhanced MR angiography of peripheral arteries in comparison to DSA in patients with peripheral artery occlusive disease

Abstract

The purpose of this study was to evaluate peripheral non-enhanced-MRA (NE-MRA) acquired with a 3D Turbo Spin Echo sequence with electrocardiographt (ECG) triggering in comparison to Digital Subtraction Angiography (DSA) as the gold standard in symptomatic peripheral artery occlusive disease (PAOD) patients. This IRB approved prospective study included 23 PAOD patients from whom three patients had to be excluded. The remaining 20 subjects were included in the analysis (15 male; mean age 62.4±15.3years). The patients first underwent DSA followed by NE-MRA on a 1.5-T whole body scanner within 24h after the DSA study. A NATIVE (Non-contrast Angiography of the Arteries and Veins) SPACE (Sampling Perfection with Application Optimized Contrast by using different flip angle Evolution) sequence at four levels (pelvis, upper leg, knee region and lower leg) was acquired. For evaluation purposes, subtracted standardized MIP (maximum intensity projection) images were generated from the NE-MRA data sets. Qualitative assessment of NE-MRA images in reference to the corresponding DSA images, as well as blinded stenosis grading of preselected segments in NE-MRA images were performed by two experienced readers. Image quality in 95 corresponding arterial segments was rated from 1 (good) to 4 (inadequate) directly comparing the NE-MRA with the corresponding DSA segment as the gold standard. Blinded stenosis grading consisted of 66 preselected stenoses rated from 1 (<10%) to 4 (>90%) in NE-MRA which were compared to the grade in the corresponding DSA. The mean image quality of NE-MRA in comparison to DSA was 2.7±1.1 (reader 1) and 3.0±1.0 (reader 2). The kappa value indicating interobserver agreement was 0.34; readers 1 and 2 rated the image quality as good in 21% and 3%, sufficient in 19% and 41%, limited in 29% and 14% and inadequate in 31% and 42%, respectively. Stenosis graduation revealed significantly higher grades in NE-MRA (reader 1: 3.0±0.7, p<0.001 and reader 2: 3.1+0.8, p<0.001) compared to DSA (mean value DSA 2.7±0.8). The kappa value indicating interobserver agreement concerning stenosis grading was 0.59. NE-MRA revealed a relatively high number of inadequate quality segments. This is in line with recently published comparable studies of the similar SPACE NE-MRA techniques. Further advance of NE-MRA techniques remains desirable for patients with PAOD.