CT of hypervascular hepatic tumors: are unenhanced scans necessary for diagnosis?

In most institutions, a combination of unenhanced and contrast-enhanced CT is used to screen patients for suspected hypervascular hepatic neoplasms. Elimination of the unenhanced examination could save time and expense and reduce the patient's exposure to radiation. We compared unenhanced and incremental bolus dynamic contrast-enhanced CT for detection of hypervascular hepatic neoplasms and evaluated the need for preliminary unenhanced hepatic CT.

We prospectively examined 101 consecutive patients with hypervascular primary hepatic malignant tumors or suspected hypervascular metastases. Primary hepatic neoplasms included hepatocellular carcinoma (seven), sarcoma (two), and hemangioendothelioma (one); extrahepatic primary tumors included breast (37) and renal cell (24) carcinoma, melanoma (15), carcinoid (nine), and endocrine and other tumors (six). All patients had both unenhanced and bolus dynamic contrast-enhanced CT with 5-mm collimated sections at 8-mm intervals. For contrast-enhanced CT, 150-180 ml of IV contrast material, a mechanical power injector, and a scanning protocol that allowed completion of liver imaging within 2 min were used. Both unenhanced and contrast-enhanced CT scans were compared for detection and conspicuousness of hepatic lesions.

Hepatic lesions were found in 34 patients. In 21 patients, all hepatic lesions seen on unenhanced scans also were apparent on contrast-enhanced scans. However, in 12 (57%) of 21 patients, lesions were more conspicuous and better defined on contrast-enhanced scans. The absolute number of lesions detected with each method of scanning differed in 12 patients. In four patients, the contrast-enhanced scan showed more lesions; in five patients, the unenhanced scan showed more lesions; and in three patients, administration of contrast material obscured some lesions shown on unenhanced scans and made others more conspicuous. If only a contrast-enhanced CT scan had been obtained, the presence of malignant hepatic neoplasm would have been missed in only one case (a patient with a single 2.8-cm metastasis from renal cell carcinoma).

Bolus dynamic contrast-enhanced CT alone correctly showed the presence or absence of primary or metastatic hypervascular hepatic tumors in 100 of 101 cases. If the goal of CT examination is detection of hypervascular hepatic lesions, use of contrast-enhanced CT alone may be adequate and the addition of unenhanced scans is not cost-effective.