Purpose: To determine the accuracy of single-kidney glomerular filtration rate (GFR) determination using contrast-enhanced dynamic magnetic resonance imaging (MRI) and the Rutland-Patlak plot technique.
Materials and methods: Twenty-eight adult patients were included. As reference method, the GFR was measured by plasma clearance using a small bolus injection of iopromide. A three-dimensional gradient-echo (GRE) sequence with a flip angle of 50 degrees was used for MRI; this showed a good linear relationship between gadolinium (Gd)-DTPA concentration and signal change when measured up to a Gd-DTPA concentration of 10 mmol/liter. A slab containing both kidneys and the abdominal aorta was measured 30 times in approximately 3.5 minutes. During this measurement, 15 mL of Gd-DTPA, 0.5 mol/liter diluted to a volume of 60 mL, was injected over 60 seconds. A Rutland-Patlak plot was calculated from the signal changes in the aorta and the renal parenchyma. Single-kidney GFR was calculated for different time windows from the Rutland-Patlak plot slope.
Results: The best correlation compared to the reference method was found with the GFR calculated from the slope of the Rutland-Patlak plot 40-110 seconds postaortic rise. Pearson's correlation coefficient was r = 0.86, SD was 14.8 mL/minute. In many of the patients, a decrease of the renal signal was observed in the excretory phase, which was probably caused by very high Gd-DTPA concentrations in the collecting tubules.
Conclusion: Single-kidney GFR can be calculated from dynamic contrast-enhanced MRI. We found a promising correlation of global GFR calculated by MRI compared to the reference method. In any future study, the amount of Gd-DTPA should by reduced to avoid artificial signal drop in the excretory phase induced by the T2* effect.
Copyright 2003 Wiley-Liss, Inc.