Liver metastases: improved detection with dynamic gadolinium-enhanced MR imaging?

Abstract

To compare dynamic gadolinium-enhanced with unenhanced magnetic resonance (MR) imaging in detection of liver metastases. Two groups of patients were prospectively examined with unenhanced and dynamic gadolinium-enhanced MR imaging. The first group (n = 48) had proved liver metastases; the second group (n = 49) did not. One set of unenhanced and one set of gadolinium-enhanced MR images were selected per patient. Three independent, blinded readers assessed the images for presence, number, location, and conspicuity of lesions. Data were analyzed with receiver operating characteristic curves, and contrast-to-noise ratios were calculated for the images. There was no statistically significant difference between the use of unenhanced and gadolinium-enhanced MR images in the differentiation of patients with from patients without metastases. The numbers of false-positive and false-negative diagnoses of individual lesions were higher (not statistically significant) with dynamic MR images than with unenhanced MR images. At dynamic MR imaging, contrast-to-noise ratio was highest in the early phase (30 seconds after injection of the contrast agent) but was not significantly different from the contrast-to noise ratio of the T2-weighted images. Dynamic gadolinium-enhanced MR imaging showed no improvement over unenhanced MR imaging in detectability of liver metastases.