Measurement of Unconjugated Estriol in Serum by Liquid Chromatography–Tandem Mass Spectrometry and Assessment of the Accuracy of Chemiluminescent Immunoassays

BACKGROUND: Unconjugated estriol (uE3) is routinely analyzed in clinical laboratories as risk assessment for Down syndrome. Immunoassays of various types are the most commonly used methods. The accuracies of RIAs and ELISAs for uE3 have been questioned, and to date there have been no independent studies investigating the accuracy of the relatively new chemiluminescent immunoassays. We developed and validated a liquid chromatography–tandem mass spectrometry (LC-MS/MS) method for uE3 measurements in serum. METHODS: Serum samples from patients in the second trimester of pregnancy were used, and uE3 concentrations were measured by LC-MS/MS and the Beckman Coulter Access® 2 and Siemens IMMULITE 2000 automatic chemiluminescent immunoassay analyzers. RESULTS: The LC-MS/MS method was validated and showed limit of detection 0.05 ng/mL; limit of quantification 0.2 ng/mL; linearity of response to 32 ng/mL; total imprecision of 16.2%, 10.4%, and 8.2% for uE3 at 1.10, 4.18, and 8.32 ng/mL, respectively; and analytical recoveries of 95.9%–104.2%. ANOVA of the correlation for LC-MS/MS results vs chemiluminescent immunoassays results showed R 2 = 0.9678 (Access 2 = 0.9305 LC-MS/MS + 0.2177, S y | x = 0.1786, P < 0.0001), and R 2 = 0.9663 (IMMULITE 2000 = 0.8849 LC-MS/MS − 0.0403, S y | x = 0.1738, P < 0.0001). Bland–Altman plots of uE3 results revealed concentration-dependent immunoassay biases. Mock risk analysis for Down syndrome showed no apparent difference in the risk assessment outcomes if the adjusted method-specific multiples of the median were used, and the assay imprecision was <10% CV. CONCLUSIONS: Standardization of immunoassay methods for uE3 analysis is needed to improve the accuracy of the measurements.