Hepatic MR imaging using ferumoxides: prospective evaluation with surgical and intraoperative sonographic confirmation in 25 cases.

Abstract

The purpose of our study was to evaluate the sensitivity and accuracy of ferumoxides-enhanced MR imaging in comparison with surgery and intraoperative sonography. We prospectively evaluated 25 consecutive studies in 24 patients who underwent ferumoxides-enhanced hepatic MR imaging before surgery and intraoperative sonography. Both 1.5-T scanners (13 cases) and 0.2-T scanners (12 cases) were used. Turbo spin-echo T2-weighted sequences were performed before and after the administration of ferumoxides and the images were compared. Lesions were classified as solid or nonsolid and tabulated on standard liver maps. The liver maps from MR imaging were compared with those from surgery and intraoperative sonography. For lesions greater than 1 cm, the regions of interest were measured and contrast-to-noise ratio was calculated. Of 93 solid lesions found at surgery, 69 were seen on unenhanced MR imaging (sensitivity, 74.2%) and 87 were seen on ferumoxides-enhanced MR imaging (sensitivity, 93.5%) (p < 0.05). Of the seven benign lesions (five cysts, two hemangiomas) found at surgery, all were correctly identified as benign on MR imaging. Two lesions identified as solid before surgery were not found at surgery. Mean lesion contrast-to-noise ratio for the unenhanced scans was 22.9 and 34.5 (p < 0.001) for the ferumoxides-enhanced scans. Subanalysis of 1.5- and 0.2-T MR imaging revealed similar results with significant (p < 0.05) increases in sensitivity for both. The average size of the lesions missed before surgery was 0.7 cm. Turbo spin-echo T2-weighted ferumoxides-enhanced MR imaging at either 1.5 or 0.2 T has value in preoperative liver assessment.