Prospective evaluation of hepatic imaging studies in the detection of colorectal metastases: correlation with surgical findings.

Nineteen patients with isolated colorectal liver metastases were preoperatively evaluated with five hepatic imaging studies. The results of these studies were correlated with operative findings. One study involved the use of EOE-13, an intravenously administered lipid contrast agent that enhances liver parenchyma on computed tomography (CT) scans (EOE-CT). Another study, arterial portography (AP-CT), was performed with injection of a water-soluble contrast agent into the superior mesenteric artery during CT scanning. Delayed CT scanning (DS-CT) was completed 4 hours after AP-CT. In addition, T1- and T2-weighted magnetic resonance imaging (MRI) (T1-MRI, T2-MRI) scans were done. All patients subsequently underwent laparotomy with identification of 78 lesions. A lesion-by-lesion analysis revealed that the sensitivity of EOE-CT (83%), AP-CT (78%), DS-CT (82%), and T1-MRI (84%) was comparable. T2-MRI had a significantly lower sensitivity (64%) compared with EOE-CT, DS-CT, and T1-MRI. The false-positive rate for AP-CT was 31%, which was significantly higher than that of all other studies. The T1-MRI examination had the lowest false-positive rate and proved to be the best hepatic imaging study in the detection of colorectal metastases. EOE-CT and DS-CT were comparable, whereas AP-CT and T2-MRI proved to be inferior tests.