[Interference in double-determinant monoclonal antibody-based assay for carcinoembryonic antigen (CEA)]

We investigated the false positive phenomena in immunoassays of carcinoembryonic antigen (CEA) using mouse monoclonal antibodies, and studied the properties of interfering substances and the method to eliminate non-specific interference. Serum samples from 2,250 patients, which indicated more than 10 ng/ml of CEA with EIA test kit (Boehringer Mannheim Yamanouchi; BMY) in about 25,000 samples of CEA tests, were subjected to extraction by heating at 70 degrees C followed by CEA measurement of the supernatant. The overall correlation was good between CEA values of the original and extraction methods, but CEA values in 55 cases were found to be remarkably higher in the non-extracted method than the extraction method. CEA titers of 20 samples of the discrepant cases were measured by various commercial kits, and the result indicated that non-specific interference was unavoidable in any kit using a monoclonal antibody-based double-determinant immunoassay, in spite of difference in degree and frequency of interference. The interfering activity of two patients serum with remarkably discrepant values eluted at the void volume of a large molecular size of more than 1,000 kDa in gel chromatography of Sephacryl S-300. The non-specific interference was reduced by addition of mouse gamma-globulin to the buffer of BMY kit and abolished by pretreatment of the serum with heat-killed cells of Staphylococcus aureus, suggesting that interference might be caused by some species of immunoglobulin being able to bind mouse gamma-globulin.(ABSTRACT TRUNCATED AT 250 WORDS)