Surveillance for hepatocellular carcinoma (HCC) in patients with recognized risk factors remains controversial. The populations for whom surveillance may be appropriate include all patients with established cirrhosis, and hepatitis B (HBV) carriers, even in the absence of cirrhosis. However, even these risk groups can be stratified into patients with higher or lower risk. The most appropriate surveillance test is periodic ultrasound examination, although the optimum screening interval has not been defined. Alphafetoprotein (AFP) is a poor surveillance test, lacking in sensitivity and specificity. There are no randomized controlled trials confirming that surveillance for HCC reduces disease-specific mortality. Modeling studies, however, have suggested that screening is cost-effective and reduces group mortality by a small amount. The criteria by which cancer surveillance programs in general can be judged have been described. Surveillance for HCC meets some of these criteria, but not all. In particular, more effective treatments have to be developed to improve the outcome of surveillance. Although there is no firm evidence to support the practice of surveillance for HCC, this has become common practice, forever preventing the definitive study from being performed. Nonetheless, surveillance is recommended in order to identify patients with small HCCs, who can be entered into trials of therapy of these tumors.
Copyright 2001 by W.B. Saunders Company.