False diagnosis and needless therapy of presumed malignant disease in women with false-positive human chorionic gonadotropin concentrations

Background 12 women were diagnosed of having postgestational choriocarcinoma on the basis of persistently positive human chorionic gonadotropin (hCG) test results in the absence of pregnancy. Most of the women had extirpative surgery or chemotherapy, or both, without significant diminution in hCG titre. Our aim was to assess whether the hCG concentrations were false-positive test results. Methods Samples were tested for hCG, hCG free β subunit, and hCG β-core fragment. Assay kinetics were also assessed, and samples were tested independently by competitive RIA. False-positive hCG concentrations were identified by two criteria: detection of hCG in serum and lack of detection of hCG and its degradation products in urine; and wide variations in results for different hCG assays. We corroborated false-positive hCG values by the lack of parallel changes in hCG results when serum was diluted, by false detection of other antigens, and by failure to detect hCG with in-house assays. Findings All 12 women met both criteria for false-positive hCG, and all had corroborating findings. In all 12 cases, a false diagnosis had been made, and most of the women had been subjected to needless surgery or chemotherapy. Assay kinetics indicated that heterophilic antibodies were responsible for the false-positive results. As a result of our findings all further therapy was stopped. Interpretation Current protocols for the diagnosis and treatment of choriocarcinoma should be modified to include a compulsory test for hCG in urine.