A Clinical Comparison of two Radioimmunoassays for Carcinoembryonic Antigen (CEA)

Publisher Website
Google Scholar
Abstract
Aliquots of the same blood samples were assayed for CEA by both the Gold and Hansen techniques. Levels of antigenemia and “positivity” vs. “negativity” were compared in 291 patients with various malignant and nonmalignant diseases. Both assays detected CEA in all disease categories studied with an overall correlation of 83-86% (p<.001). Apparent “differences” between the two assays were primarily quantitative. Hansen (Nutley) values were slightly higher than Hansen (Boston) values and both were slightly higher than Gold values in the low ranges of antigenemia. The Gold assay tended to read higher than either Hansen assay in the group with alcoholic liver disease, when values (Gold) were greater than 4 ng/ml. The lowest frequency of agreement between assays occurred in those groups with low-levels of antigen such as the postoperative colon carcinoma and the inflammatory bowel disease patients. This reflects variations inherent in the assays as currently performed. Variation of ± 1 around the 2.5 ng/ml “cut-off” between “positive” and “negative” accounted for 83% of the apparent noncorrelations. Hence, clinical interpretations should not be based on a sharp “cut-off” level at the present time. Standardization of assay media and redefinition of “positivity” in individual laboratories will be necessary before clinically significant differences can be ascertained. Neither test is specific for carcinoma and decisions regarding treatment should not be based on this test alone without consideration of all appropriate clinical, laboratory and radiologic examinations. Further studies of the clinical potentials of the CEA assays are warranted.