Although alpha-fetoprotein (AFP) is regarded as the reference marker for hepatocellular carcinoma (HCC), it sometimes produces false results. The objective of this study was to see if some of the readily available laboratory markers could complement AFP to improve the laboratory diagnosis of HCC. The markers tested and their sensitivities were: CA 125, 92%; ferritin, 71.3%; CA 19-9, 69.8%; beta-2-microglobulin (B2M), 53.3%; CA 72-4, 13.6%; and carcinoembryonic antigen (CEA), 10.6%. In comparison, AFP had a sensitivity of 58.8%. CA 72-4 and CEA (at the "tumour" cut-off level of 20 ng/ml) had specificities of 100%, and AFP, 97.4%. The specificities of the other markers were less impressive: CEA, 77.8% (at the cut-off level of 5 ng/ml); ferritin, 48.6%; CA 125, 48.5%; B2M, 39.6%; and CA 19-9, 37.3%. The efficiencies of the markers for HCC, which are based on the consideration of sensitivity and specificity together, were as follows: AFP, 77.6%; CA 125, 71.3%; ferritin, 60.5%; CA 19-9, 55.3; B2M, 46.9%; CEA, 40.8%; and CA 72-4, 34.5%. The receiver-operating characteristic plots confirmed AFP to be the most efficient marker for HCC. Nevertheless, it is proposed that CA 125 be combined with AFP for HCC screening because of their excellent sensitivity and specificity, respectively: a negative result for both, or even just CA 125 alone, would indicate that the disease is unlikely while a positive AFP (which would likely occur with a positive CA 125) would make its presence highly probable. A positive CA 125 and negative AFP would be equivocal for HCC. Other markers in combination with AFP are less useful.
* Title and MeSH Headings from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.