Serum diagnostic tests for pancreatic cancer.

Serological tests for pancreatic cancer have been criticized too harshly as a result of being tested in inappropriate roles. They are never likely to become sufficiently specific for screening an asymptomatic population unless some way can be found of defining a population with a particularly high risk for the disease. Nor are they appropriate in the investigation of the jaundiced patient. The markers that are carried by secreted mucins seem the most promising and in view of the marked heterogeneity of carbohydrate expression on mucins a combination of tests for two or three carefully selected markers is likely to be better than one. The high cost per test that results from using commercially available radioimmunoassay kits with a short shelf-life can be reduced by using enzyme-linked assays which have a much longer shelf-life. These tests are likely to be of most help in the investigation of non-jaundiced patients with unexplained abdominal pain or weight loss. In this group of patients it seems probable that serological tests will compliment scanning techniques but further studies are needed to assess this.