Between-method variation in human chorionic gonadotropin test results.

Abstract

Results on sera and calibrators vary 1.4- to 2.3-fold among commercial human chorionic gonadotropin (hCG) assays. The relative contributions of calibrators, standards, hCG charge isoforms, and major structural variants to this variation have not been quantified. Purified hCG was separated by isoelectric focusing into four fractions with pI ranges of 3-4, 4-5, 5-6, and 6-7. These four fractions together with pure hCG, hyperglycosylated hCG, hCG beta-subunit (hCGb), nicked hCG, and hCGb core fragment (hCGbcf) were tested in nine commonly used commercial serum assays for hCG. The compositions of pure hCG preparations, standards, and commercial hCG preparations were determined by immunoassay. The three pure hCG preparations and the four hCG charge isoforms each showed parallel responses in the nine commercial hCG assays. Although wide variations were found in the detection of hCG structural variants by the nine assays (range for hyperglycosylated hCG, 468-1544 IU/L; for hCGb, 3187-5535 IU/L; for nicked hCG, 2736-4240 IU/L; and for hCGbcf, <2-130 IU/L), this did not correlate with the between-method variation observed in results for the three pure hCG preparations. Commercial preparations of hCG and calibrators showed great variation in their content of hCG structural variants (from 34% to 100% intact hCG). Intermethod differences in hCG results were not explained by changes in responses attributable to hCG charge isoforms or to hCG structural variants, but wide variation was observed in concentrations of hCG structural variants in calibrators and in detection of these structural variants. Differences in assay specificity and in composition of the calibrators are the most likely sources of between-method variation.