Clinical use of tumor markers in oncology.

The perfect tumor marker would be one that was produced solely by a tumor and secreted in measurable amounts into body fluids, it should be present only in the presence of cancer, it should identify cancer before it has spread beyond a localized site (i.e., be useful in screening), its quantitative amount in bodily fluids should reflect the bulk of tumor, and the level of the marker should reflect responses to treatment and progressive disease. Unfortunately, no such marker currently exists, although a number of useful but imperfect markers are available. The predominant contemporary markers are discussed here by chemical class, as follows: glycoprotein markers, including carcinoembryonic antigen (CEA), alpha-fetoprotein (AFP), beta-human chorionic gonadotropin (beta-hCG), and prostate specific antigen (PSA); mucinous glycoproteins, including CA 15-3, CA 19-9, mucinous-like cancer antigen and associated antigens, and CA 125; enzymes, including prostatic acid phosphatase (PAP), neuron specific enolase (NSE), lactic acid dehydrogenase (LDH), and placental alkaline phosphatase (PLAP); hormones and related endocrine molecules, including calcitonin, thyroglobulin, and catecholamines; and, molecules of the immune system, including immunoglobulins and beta-2-microglobulin. The biologic properties of each group of tumor markers are discussed, along with our assessment of their role in clinical medicine today.