The purpose of this study was to compare the efficacy of unenhanced and
ferumoxides-enhanced magnetic resonance imaging with that of dual-phase spiral CT and spiral CT during arterial portography (CTAP) for the detection of colorectal liver
metastases. Fourteen patients with liver
metastases candidates for partial
hepatectomy were examined with dual-phase spiral CT, unenhanced and
ferumoxides-enhanced MR imaging at 1.5 T, and spiral CTAP. Imaging tests were read blinded, prospectively, quantitating number of lesions excepting CTAP which used US to exclude
cysts. Subsequent intraoperative US and pathologic findings were correlated with preoperative imaging results. At surgery, 36 lesions 0.5-13 cm in diameter (mean+/-standard deviation, 2.9+/-2.1 cm) were identified. Dual-phase spiral CT depicted 21/36 (58%); precontrast MR imaging, 19/36 (53%);
ferumoxides-enhanced MR imaging, 30/36 (83%); and spiral CTAP, 34/36 (94%) lesions.
Ferumoxides-enhanced MR imaging was significantly more sensitive than spiral CT and unenhanced MR imaging (P < 0.01). The difference in sensitivity between
ferumoxides-enhanced MR imaging and spiral CTAP was not statistically significant (P> 0.1). Spiral CTAP, however, depicted nine false-positive lesions (2
hemangiomas, 7 perfusion defects). The positive predictive value was 79% for spiral CTAP and 100% for combined pre- and postcontrast MR imaging. We conclude that
ferumoxides-enhanced MR imaging is superior to unenhanced MR imaging and biphasic spiral CT for depiction of colorectal liver
metastases. Further investigation is needed to clarify whether MR imaging with use of
ferumoxides might replace spiral CTAP for preoperative evaluation of liver resection candidates.