PURPOSE: To compare the diagnostic performance of gadoxetic acid-enhanced MRI with ferucarbotran-enhanced MRI for the detection of liver metastases. MATERIALS AND METHODS: Thirty-six patients with 80 liver metastases who underwent gadoxetic acid-enhanced MRI using a three-dimensional volumetric interpolated technique and ferucarbotran-enhanced MRI with a mean interval of 7 days (range, 5-10 days) were included in this study. Two observers independently interpreted the two sets of images - the gadoxetic acid set (unenhanced, early dynamic and 20min delayed phase images) and the ferucarbotran set (unenhanced and ferucarbotran-enhanced T2*-weighted-gradient echo and T2-weighted turbo spin echo images). Diagnostic accuracy was evaluated using the alternative-free response receiver operator characteristic (ROC) method. Sensitivity and positive predictive value were also evaluated. RESULTS: There was a trend toward increased areas under the ROC curve (Az values) for the gadoxetic acid set (0.950, 0.948) as compared with the ferucarbotran set (0.941 and 0.939) of images, but no significant difference was found for both observers (p<0.05). Sensitivities of the gadoxetic acid set (93.8% and 92.5%) were also slightly better than those of the ferucarbotran set (88.8% and 87.5%) with no significant difference (p=0.13). The two image sets showed similar positive predictive values (98.7% and 98.6%, respectively). CONCLUSIONS: Gadoxetic acid-enhanced MRI showed comparable diagnostic performance to ferucarbotran-enhanced MRI for the detection of liver metastases. Copyright (c) 2009 Elsevier Ireland Ltd. All rights reserved.
PURPOSE: To compare the diagnostic performance of gadoxetic acid-enhanced MRI with ferucarbotran-enhanced MRI for the detection of liver metastases. MATERIALS AND METHODS: Thirty-six patients with 80 liver metastases who underwent gadoxetic acid-enhanced MRI using a three-dimensional volumetric interpolated technique and ferucarbotran-enhanced MRI with a mean interval of 7 days (range, 5-10 days) were included in this study. Two observers independently interpreted the two sets of images - the gadoxetic acid set (unenhanced, early dynamic and 20min delayed phase images) and the ferucarbotran set (unenhanced and ferucarbotran-enhanced T2*-weighted-gradient echo and T2-weighted turbo spin echo images). Diagnostic accuracy was evaluated using the alternative-free response receiver operator characteristic (ROC) method. Sensitivity and positive predictive value were also evaluated. RESULTS: There was a trend toward increased areas under the ROC curve (Az values) for the gadoxetic acid set (0.950, 0.948) as compared with the ferucarbotran set (0.941 and 0.939) of images, but no significant difference was found for both observers (p<0.05). Sensitivities of the gadoxetic acid set (93.8% and 92.5%) were also slightly better than those of the ferucarbotran set (88.8% and 87.5%) with no significant difference (p=0.13). The two image sets showed similar positive predictive values (98.7% and 98.6%, respectively). CONCLUSIONS:Gadoxetic acid-enhanced MRI showed comparable diagnostic performance to ferucarbotran-enhanced MRI for the detection of liver metastases. Copyright (c) 2009 Elsevier Ireland Ltd. All rights reserved.