Prediction of Split Renal Function Improvement After Renal Artery Stenting by Blood Oxygen Level–Dependent Magnetic Resonance Imaging

BACKGROUND: The discrepancy between the high technical success rate and the relatively low clinical response rate of renal artery stenting (RAS) raises the importance to identify atherosclerotic renal artery stenosis (ARAS) patients who are most likely to benefit from RAS. This study aimed to investigate the feasibility and accuracy of blood oxygen level-dependent magnetic resonance imaging (BOLD-MRI) in predicting split renal function (SRF) improvement after RAS in patients with ARAS. METHODS: Thirty patients with severe ARAS who were treated with RAS were enrolled. Baseline cortical and medullary R2* values of each kidney were measured by BOLD-MRI, and each patient’s SRF was evaluated by nuclear renal dynamic imaging at baseline and 1-month follow-up. RESULTS: In total, 35 severe stenotic renal arteries of the 30 patients were analyzed. At 1-month follow-up, 34 kidneys (97.1%) of severe ARAS had acquired SRF. SRF improved in 12 kidneys of 10 patients. The cortical R2* and medullary R2* values in the SRF improvement kidneys were higher than those in the non-improvement kidneys (P ≤ 0.001). The area under the curve of medullary R2* was 0.879 (95% confidence interval [CI] 0.736–1.000). A medullary R2* value ≥29.1 s(–1) was noted to provide good sensitivity (0.833, 95% CI 0.552–0.970) and specificity (0.864, 95% CI 0.667–0.953) in predicting SRF improvement. Medullary R2* value was the only independent predictor of SRF improvement in multivariable analysis (P = 0.034, OR 3.017, 95%CI 1.089–8.358). CONCLUSION: This study showed that a BOLD-MRI medullary R2* value ≥29.1 s(–1) was an excellent predictor of SRF improvement in patients with severe ARAS who underwent renal artery stenting.