Comparison of ferumoxytol‐enhanced MRA with conventional angiography for assessment of severity of transplant renal artery stenosis

Purpose To determine the accuracy of ferumoxytol‐enhanced magnetic resonance angiography (MRA) in assessing the severity of transplant renal artery stenosis (TRAS), using digital subtraction angiography (DSA) as the reference standard. Materials and Methods Our Institutional Review Board approved this retrospective, Health Insurance Portability and Accountability Act‐compliant study. Thirty‐three patients with documented clinical suspicion for TRAS (elevated serum creatinine, refractory hypertension, edema, and/or audible bruit) and/or concerning sonographic findings (elevated renal artery velocity and/or intraparenchymal parvus tardus waveforms) underwent a 1.5T MRA with ferumoxytol prior to DSA. All DSAs were independently reviewed by an interventional radiologist and served as the reference standard. The MRAs were reviewed by three readers who were blinded to the ultrasound and DSA findings for the presence and severity of TRAS. Sensitivity, specificity, and accuracy for identifying substantial stenoses (>50%) were determined. Intraclass correlation coefficients (ICCs) were calculated among readers. Mean differences between the percent stenosis from each MRA reader and DSA were calculated. Results On DSA, a total of 42 stenoses were identified in the 33 patients. The sensitivity, specificity, and accuracy of MRA in detecting substantial stenoses were 100%, 75–87.5%, and 95.2–97.6%, respectively, among the readers. There was excellent agreement among readers as to the percent stenosis (ICC = 0.82). MRA overestimated the degree of stenosis by 3.9–9.6% compared to DSA. Conclusion Ferumoxytol‐enhanced MRA provides high sensitivity, specificity, and accuracy for determining the severity of TRAS. Our results suggest that it can potentially be used as a noninvasive examination following ultrasound to reduce the number of unnecessary conventional angiograms. Level of Evidence: 3 J. Magn. Reson. Imaging 2017;45:779–785.