undetectable by computed tomography (CT) scan. We are not given this information in the first study, but benign regenerative nodules are typically not visible by CT. If a focal lesion is confirmed by CT scan, again the question of malignancy should be raised. (d) Most lesions in the first study but not the second study enlarged with follow-up, which should also suggest HCC.
So what should we do with the asymptomatic patient with known cirrhosis and a focal lesion demonstrated by ultrasound? Most are going to have HCC. If the a-fetoprotein (AFP) level is high ( > 1,000 ng/ml), the patient almost certainly has HCC, and histological confirmation may not be necessary. A low AFP does not rule out malignancy. Although CT scan is not feasible for screening programs in the Far East, in the United States most ultrasound lesions will be confirmed by CT before obtaining the histological study. Ultrasound-guided, FNA should then be performed and, if histologically benign, repeated with a thin liver-biopsy needle. If the diagnosis of HCC is not made, the prudent course is to observe the patient. A rising AFP level and/or sonographic enlargement of the lesion should prompt further invasive evaluation by either an ultrasound-guided or laparoscopic biopsy procedure.
Are regenerative nodules premalignant lesions? The cirrhotic liver per se is associated with an increased risk of HCC, the risk varying with the cause of the liver disease. However, regenerative nodule is not synonymous with focal liver lesion. Every focal mass lesion in a cirrhotic liver should simply be considered malignant until proven otherwise (e.g., with a follow-up of more than 2 yr without observed progression). With careful evaluation, most will be shown to be malignant initially.
The evidence that large regenerative nodules have greater malignant potential than the surrounding cirrhotic liver is slim at best.