No evidence of the hook effect in peritoneal fluid CA 125 measurement using immunoenzymatic second generation assays: comparison with immunoradiometric assays.

OBJECTIVE: To monitor the occurrence of the hook effect in measurements of peritoneal fluid CA 125 levels using two different immunoenzymatic second generation assays (ETI-II and EIA-II), and to compare these results with those obtained using the respective immunoradiometric versions of the assays (IRMA-II and ELSA-II). STUDY DESIGN: CA 125 levels were determined in peritoneal fluid and serum samples obtained from 45 women with gynecological diseases. The assays were carried out using IRMA-II and ETI-II (Sorin Biomedica) and ELSA-II and EIA-II (CIS Bio International) assays. Occurrence of the hook effect and linearity of the assays were evaluated. Statistical analyses were performed by Wilcoxon's test and linear regression analysis. RESULTS: Undiluted peritoneal fluids, assayed for their CA 125 content, showed falsely underestimated values of the antigen when IRMA-II and ELSA-II assays were performed. The phenomenon disappeared only at high dilutions of the sample (> 50). Conversely, immunoenzymatic ETI-II and EIA-II assays performed on undiluted peritoneal fluids did not show underestimated CA 125 values. CA 125 values obtained by immunoenzymatic assay were lower than those obtained using their respective immunoradiometric versions at a dilution of 1:100 (P < 0.001). A good correlation was observed between ELSA-II and EIA-II (r = 0.929) CA 125 values. CONCLUSION: The EIA-II immunoenzymatic assay appears to be more suitable for CA 125 measurement in peritoneal fluid in that it is not subject to the hook effect and its results correlated well with those obtained via its immunoradiometric version.