ECE2020 Audio ePoster Presentations Reproductive and Developmental Endocrinology (79 abstracts)
University hospital , Bordeaux, France
Anti-Mullerian Hormone (AMH) is a 140 kDa gonadal glycoprotein of the TGFβ family, with two subunits linked by two disulfide bridges. Since a first assay in 1990, three generations of assays occurred: the 1st and 2nd were manual assays. The 3rd was automated: Access AMH Kits (Beckman Coulter) and Elecsys AMH (Roche). We had the opportunity to compare two new analysers with the Elecsys AMH: Vidas (Biomérieux) and Lumipulse G (Fujirebio). The later are one-step sandwich immunoenzymatic (fluorescence detection) and two-step sandwich chemiluminescence, respectively. The population studied consisted of 50 women of reproductive age (27 [20.5–37.7] yr, median [5th-95th] percentiles; BMI 21.7 [18.2–29.1] kg/m²). The sera collected were centrifuged, aliquoted in 3 aliquots and frozen until the day of the assay. Intra-laboratory precisions reported by the suppliers were: CV < 2.1% for sample panels with concentrations from 0.41 to 22.44 ng/ml for the Lumipulse and CV < 10.6% for sample panels with concentrations from 0.22 to 7.37 ng/ml for the Vidas. The results obtained were closely correlated with no significant deviation from linearity (Passing Bablok test): Elecsys vs Vidas correlation coefficient r = 0.976, [Elecsys] = 0.0778767 + 1.086758*[Vidas] ng/ml & Elecsys vs Lumipulse r = 0.913, [Lumipulse] = 0.0308871 + 1.097997*[Vidas] ng/ml. On the Bland-Altman graphs, 3 outliers were found outside the concordance limits for the Vidas against 1 outlier for the Lumipulse. Despite these very good correlations and minor biases, an important problem rests with the proposed reference values which are not similar. For instance, for a 30-34 yr old woman: Elecsys 2.81 [0.71–7.59], Lumipulse 1.557 [0.065–10.042], Vidas 3.55 [1.19–7.00] ng/ml. A combination of reasons may explain this. Firstly, an international AMH standard is not yet used for theses assays. Furthermore, AMH concentrations can be influenced by different factors (ethnicity, tobacco, oral contraceptives, etc.). Reference values may thus be different from so-called normal values in healthy subjects. Specifics about the determination of reference ranges by the manufacturers are then critical (Ferguson, Reprod Biomed Online, 2018). Indeed, precise identification of these ranges is mandatory to assess the usefulness of determining AMH concentration with any apparatus.