Falsely Low Results Obtained with the Hybrltech Tandem # { 174

Prostate-specific antigen (PSA) is a recently described marker for prostate carcinoma (1). In patients with prostatic cancer the concentration of PSA in serum can be as much as 10 mg/L, whereas concentrations in healthy men are <2.8 zg/L (2). Hybritech’s immunoradiometric assay for PSA has been in routine use in our laboratory for 15 months. Its reproducibility is generally good, but two samples with values of 48 and 38 p.g/L showed abnormal variation between duplicates. We re-assayed these samples diluted 10to 1000-fold, and found that their PSA concentrations actually were 8.60 and 8.90 mg/L, respectively. This suggested that a “hook effect”-i.e., a decreased response in samples containing a large excess of antigen-had caused a falsely low result. The hook effect is a phenomenon usually associated with one-step “sandwich”-type assays (3). Therefore, we modified the PSA assay to a two-step assay with 16+2 h incubation. Surprisingly, an identical hook effect was also observed in this procedure. Fractionation of the samples by gel chromatography (Sephacryl S-200) showed that the main PSA activity was eluted in a volume corresponding to a molecular mass of -30 kDa, which corresponds to that of PSA (1). A smaller 16-kDa PSA species contributed -9% of the total PSA activity. When we assayed these components at various dilutions, the dilution curves were parallel. A large difference in affinity for the solid-phase antibody of either of these components could have explained the hook effect. Another potential cause, heterogeneity of the solid-phase antibody (3), is unlikely, because both antibodies used in the assay were monoclonal. Since detecting falsely low results in the first two patients, we assayed 320 samples at two dilutions and found four more samples with true values of -3.00 mg/L that, undiluted, gave falsely low results of -400 g/L. This problem has to be considered when the PSA assay is used to monitor patients with prostatic cancer.