Evaluation of malignant liver tumors: biphasic MS-CT versus quantitative contrast harmonic imaging ultrasound.

摘要

PURPOSE: It was the aim of our study to establish the extent to which contrast enhancement with SonoVue(R) in combination with quantitative evaluation of contrast-medium dynamics facilitates the detection of malignant hepatic tumors. MATERIALS AND METHODS: One hundred patients with histologically confirmed malignant or benign hepatic tumors (maximum size 5 cm) were analyzed. We used a high-end ultrasound machine (Logic 9, GE Healthcare, Milwaukee, WI, USA), with a multifrequency curved array transducer (2.5 - 4 MHz), Contrast-enhanced ultrasound (bolus injection 2.4 mL SonoVue(R)) was carried out with the intermittent breath-holding technique. Native vascularization was analyzed with power Doppler sonography. The contrast-enhanced dynamic ultrasound investigation was carried out with contrast harmonic imaging in the true detection mode during the arterial, portal venous and late phases. The mechanical index was set at 0.15. Perfusion analysis was performed by post-processing of the raw data (time intensity curve [TIC] analysis). Biphasic 16- or 64-slice multislice computed tomography served as reference method in nearly all cases. RESULTS: One hundred patients with 59 malignant (43 colon, 5 breast, 2 endocrine metastases, 7 hepatocellular carcinomas and 2 kidney cancers) and 41 benign (12 circumscribed fatty changes, 2 abscesses, 7 focal nodular hyperplasias, 5 complicated cysts and 15 hemangiomas) tumors were included. The CT classification was true positive in 71 of 92 patients, false negative in 8 cases, and in 13 cases no final diagnosis was possible; sensitivity was 96.7 % and specificity was 71.4 % for CT. The quantitative contrast harmonic imaging ultrasound classification was true positive in 98 of 100 patients and false negative in 2 cases; the sensitivity was 98.6 % and the specificity was 96.6 %. The Fisher test showed a significant difference at p < 0.05. No investigator-dependency was noted. CONCLUSION: In our study contrast-enhanced ultrasound was more accurate than multislice computed tomography in the prediction of malignancy and benignity of liver tumors.