Objective: The purpose of this study was to determine the sensitivity of portal-phase IV contrast-enhanced helical CT in detecting malignant hepatic tumors using a lesion-by-lesion comparison with surgical and pathologic findings.
Materials and methods: Preoperative portal-phase contrast-enhanced helical CT studies in 21 patients (13 men, eight women) with malignant neoplasms of the liver (12 patients with primary hepatic tumors, nine patients with metastases) were reviewed retrospectively by three radiologists with the knowledge that hepatic malignancies were present in each case. Helical CT findings were compared with intraoperative findings by palpation (21 patients), sonography (15 patients), and the results of pathologic examination of resected specimens (10 patients) in a lesion-by-lesion manner. Sixty-four malignant nodules were identified in resected specimens or at surgery by palpation or intraoperative sonography. For each nodule identified by surgical or pathologic means, CT findings were reviewed for the presence of a lesion of similar size and location.
Results: The readers identified 52 of 64 nodules on helical CT scans for an overall sensitivity of 81%. Sixteen of 21 primary tumors (76%) and 36 of 43 metastatic tumors (84%) were identified on CT scans. The sensitivity was 91% for nodules greater than 1 cm (42 of 46) and 56% for nodules 1 cm or less (10 of 18). False-positive findings occurred in 4% of lesions seen on helical CT scans.
Conclusion: Our results show that portal-phase contrast-enhanced helical CT is extremely sensitive (91%) for detecting malignant hepatic tumors greater than 1 cm, is relatively insensitive (56%) for tumors less than 1 cm, and has a low false-positive rate. These results compare favorably with published results for CT during arterial portography, the current gold standard for liver tumor detection, suggesting that portal-phase IV contrast-enhanced helical CT is an excellent noninvasive preoperative study prior to hepatic resection or cryosurgery.