ESPE Abstracts (2018) 89 RFC10.4

Comparative Analysis between Immunoassay and Tandem Mass Spectrometry for Androgens before and after Human Recombinant Gonadotrophin in Children with Genital Ambiguity and 46,XY Karyotype

Letícia Oliveiraa, Gil Guerra-Juniora, Carlos Longuib, Guilherme Guaragna-Filhoa, José Luiz Costaa, Rafael Lanaroa, David Silvaa, Maricilda Melloa, Andrea Maciel-Guerraa & Andre Morcilloa


aUNICAMP (Universisty of Campinas), Campinas, Brazil; bFCMSCSP (Medical Science College at Santa Casa in São Paulo), São Paulo, Brazil


Liquid chromatography associated with tandem mass spectrometry (LC-MS/MS) is currently considered the gold standard for steroid measurement. The aim of this study was to compare traditional immunoassay and LC-MS/MS methods for androgens measurement before and after human recombinant chorionic gonadotrophin (hrCG) stimulation in children with diagnosis of disorder of sex development (DSD) with 46,XY karyotype and past of normal testosterone secretion. We evaluated 19 patients, five cases of partial androgen insensitivity syndrome (PAIS), four of 5α-reductase type 2 deficiency and 10 of idiopathic 46,XY DSD, all prepubertal and non-gonadectomized, before and seven days after application of 6,500 IU of hrCG (Ovidrel®). Total testosterone, dehydroepiandrosterone (DHEA) and androstenedione were measured by immunoassay and LC-MS/MS. The correlation between the tests was analyzed by the Intraclass Correlation Coefficient (ICC) and Spearman Correlation Coefficient (SCC) tests, and for the concordance analysis, the Passing & Bablok (PB) regression and the Bland & Altman (BA) method. Regarding the ICC and SCC coefficients, respectively, the total testosterone showed an excellent correlation in both (0.958 and 0.964), moderate DHEA in both (0.562 and 0.716) and androstenedione a poor correlation in the ICC (0.363) and moderate in the SCC (0.735). By the PB method, the three androgens presented linear relationship, however the androstenedione presented proportional and systematic errors, the testosterone systematic errors, and the DHEA none of these errors. By the BA method there is a tendency of the immunoassay towards LC-MS/MS to overestimate the values of testosterone and androstenedione and to underestimate those of DHEA. In conclusion, the LC-MS/MS method and immunoassays, despite having good correlation and being linearly related, systematic and/or proportional errors were detected, with the immunoassay overestimating the testosterone and androstenedione values and underestimating those of DHEA in relation to LC-MS/MS.