Intraoperative Ultrasound of the Liver in Primary and Secondary Hepatic Malignancies: Comparison With Preoperative 1.5-T MRI and 64-MDCT

Abstract

The purpose of this study is to compare intraoperative ultrasound and preoperative contrast-enhanced MRI or 64-MDCT for the depiction of malignant lesions and for prediction of hepatic segments positive and negative for malignancy in patients undergoing partial hepatic resection. In this retrospective study, 292 patients undergoing hepatic resection for metastatic colorectal cancer (n = 168), hepatocellular carcinoma (n = 70), or other hepatic malignancies (n = 54) were included. The sensitivity and negative predictive value of intraoperative ultrasound and preoperative cross-sectional imaging were calculated. The mean (± SD) time intervals to surgery were 37.6 ± 26 days for 64-MDCT and 48.1 ± 34 days for MRI. Surgical histopathologic examination was the reference standard. Changes in surgical management were recorded. Logistic regression models were used to estimate and compare proportions. For all 561 malignant lesions, the sensitivity of intraoperative ultrasound was 95.1%, compared with 96.8% for 64-MDCT (p = 0.025) and 94.4% for MRI (p = 0.960); 64-MDCT was also more sensitive than intraoperative ultrasound in identifying positive liver segments (p = 0.013). After controlling for patient group and time interval between imaging and surgery, the negative predictive value of 64-MDCT and MRI was higher than that of intraoperative ultrasound (p < 0.001 and p = 0.040, respectively). In only eight cases (2.7%) was surgical management changed after intraoperative ultrasound. For patients undergoing partial liver resection for hepatic malignancies, 64-MDCT and MRI have an equivalent or higher sensitivity in identifying hepatic segments with malignancy, and both 64-MDCT and MRI appear to have a higher predictive value for identifying disease-free segments than does intraoperative ultrasound.