A Study on Predicting Split Renal Function Using Renal Blood Flow Weighted Estimated Glomerular Filtration Rate From Arterial Spin Labeling <scp>MRI</scp>

BACKGROUND: Traditional methods for evaluating split renal function, including contrast-enhanced CT and radionuclide imaging, have limitations related to ionizing radiation or potentially nephrotoxic contrast agents. PURPOSE: To investigate the value of renal blood flow (RBF)-weighted estimated glomerular filtration rate (eGFR) derived from arterial spin labeling (ASL)-MRI in predicting split renal function in patients with suspected renal artery stenosis (RAS). STUDY TYPE: Retrospective. POPULATION: Fifty-three consecutive patients (21 females; mean age, 61.28 ± 13.50 years) with suspected RAS and renal insufficiency underwent renal ASL-MRI. FIELD STRENGTH/SEQUENCE: 1.5T; 3D pseudo-continuous ASL technique. ASSESSMENT: RBF and regions of interest (ROI) areas of the renal cortex and entire kidney were measured on oblique coronal ASL-MRI perfusion maps. The RBF-weighted eGFR (RBFw eGFR) for each kidney was derived by normalizing the RBF-area product to the bilateral total eGFR. The radionuclide GFR (rGFR) obtained from renal dynamic radionuclide imaging was used as the reference standard. STATISTICAL TEST: Paired t-tests, one-way repeated-measures ANOVA, Bonferroni test, Pearson's correlation analysis, and receiver operating characteristic (ROC) curves were used. Statistical significance was set at p = 0.05. RESULTS: The cortical and entire RBFw eGFR showed strong significant correlations with rGFR (r: 0.799 vs. 0.800). The cortical and entire RBFw eGFR had high diagnostic efficiency for differentiating severe, moderate, and mild renal insufficiency (AUC: 0.934 vs. 0.936, 0.927 vs. 0.928, and 0.873 vs. 0.865). DATA CONCLUSIONS: The RBFw eGFR derived from ASL-MRI demonstrated a significant correlation with rGFR and high diagnostic performance for differentiating split renal dysfunction grades, suggesting potential for split renal function assessment in patients with RAS. EVIDENCE LEVEL: 3 (Nonconsecutive cohort studies, usually retrospective with an imperfectly applied reference standard). TECHNICAL EFFICACY: 2 (diagnostic accuracy/clinical efficacy).