Détermination de la testostérone biodisponible chez la femme adulte par chromatographie en phase gazeuse-spectrométrie de masse avec dilution isotopique (ID-GCMS). Comparaison avec la méthode de routine à  la testostérone tritiée

Accurate measurement of androgens is particularly problematic in women who have lower plasma concentrations of total T (T tot) and Bioavailable Testosterone (T Bio) respectively 10-fold and 100-fold lower than men. Moreover, women taking oral estrogens have higher levels of T tot because of the increase in sex hormone binding globulin (SHBG) levels. It results that Bioavailable or Free T is decreased and this is the best parameter to measure because the only non-SHBG bound Testosterone (Free T + albumin-bound T) diffuses easily from circulation to tissues and is available for target cells. Direct RIA using the analog tracer technique has little to recommend it, salivary assays are highly variable and those that directly measure Free T in the dialysate are not commercially available and are labor intensive. The objective of this study is to evaluate a sensitive method for the measurement of the low total and bioavailable Testosterone levels in Female Androgen Insufficiency (FAI) other than menopause with androgen values at or below the lowest quartile of the reference range for reproductive age women (20-40 years) in conjunction with the presence of clinical symptoms and adequate estrogen status. An application of the most specific technique of steroid analysis, the Stable Isotope Dilution/Gas Chromatography-Mass Spectrometry (ID-GCMS) to validate Bioavailable Testosterone assays has been lacking to date. Therefore, we compare two experimental approaches (indirect and direct) to determine T Bio: 1) the indirect determination by measurement of the percentage of non-SHBG bound steroid after incubating at 37 °C the serum samples with freshly purified tracer dose of tritiated testosterone [3H]T and precipitation with saturated ammonium sulfate solution (final: 50% saturation); to obtain the concentration of T Bio: T Bio 3H E170 (III) or T Bio 3H MS (IV), we multiplied this percentage by total Testosterone: T tot E170 (I) or T tot MS (II) determined respectively by the ECLIA method on Modular E-170 Roche and by ID-GCMS; 2) the direct measurement by ID-GCMS of the T Bio MS (V) concentration in the supernatant after the ammonium sulfate precipitation step. Sensitivity of ID-GCMS (with a signal to noise of 3:1) is 2 pg injected of derivatized Testosterone. For the ID-GCMS methods II and V, the overall method recoveries are 82% and 92% respectively, CV within-assay are 3.9% (126 pg/ml) and 6.2% (13.8 pg/ml), CV between-assays are 4.3% (126 pg/ml) and 17.3% (15.4 pg/ml). For the T Bio 3H methods III and IV, within- and between-run CV are 4.5% at 2.8 pg/ml and 3.7% at 58.4 pg/ml respectively. Our data show highly significant (P < 0.001) correlation coefficients for Ttot: I with II (r = 0.88) and between indirect (III or IV) and direct techniques (V) for T Bio : III with V (r = 0.85) and IV with V (r = 0.90). This demonstrates the analytical validity of the indirect approach if we consider the direct technique by ID-GCMS as a reference measurement procedure. The method T tot E170 (I) has acceptable performance (between-run CV = 20% at 120 pg/ml) for routine investigations but T tot MS (II) is required at lower levels (between-run CV = 4.3% at 126 pg/ml) or whenever maximal reliability is requested. The validity of the T Bio 3H results depends on the validity of the total T results on which they are calculated. Despite this limitation, if we use the criteria of maintaining imprecision within- and between-run (CV) < 5% et < 10% respectively at or below (T Bio < 6 pg/ml) the lowest quartile of the reference range for reproductive age women (20-40 years), T Bio 3H methods III and IV are the only adequate techniques for routine use although the direct ID-GCMS method V has high but insufficient sensitivity.