Heterophilic antibodies in human sera causing falsely increased results in the CA 125 immunofluorometric assay.

An in-house OC 125 monoclonal antibody-based "sandwich" immunofluorometric assay (IFMA) described previously (Clin Chem 1987; 33:2191-4) gave higher results for CA 125 in 37 of 123 serum samples than did a commercially available immunoradiometric assay (IRMA). Discordant results between the two assays became concordant when measurements of the samples were repeated with normal mouse serum (100 mL/L) included in the IFMA reagent. The murine immunoglobulins are thought to block the ability of the heterophilic antibodies in the human serum samples to cross-link the labeled antibody with the solid-phase antibody. Using an enzyme immunoassay, we demonstrated human anti-mouse antibodies (HAMA) in most of the discordant samples examined. We tested the ability of nonimmune sera from other animal species to lower the apparent CA 125 concentrations of the spurious samples and observed that rat, goat, and sheep serum were less effective than mouse serum. One serum sample was discovered to give a falsely increased CA 125 result with the IRMA, but this increase could be prevented by adding murine serum to the IRMA reagent. We conclude that falsely increased CA 125 results are best prevented by adding murine serum (or murine antibodies) to the assay buffer.