The access to a sensitive, specific and reproducible assay for the measurement of androgen levels is challenging, particularly in the case of low sample concentrations and in the presence of other steroids with closely related structures which may cross-react the assay (Rosner et al., 2010). For the past five decades, immunoassay has been the routine assessment for androgen concentrations, and for steroid hormones in general. A great deal of our knowledge about the role of sex steroid hormones in normal cycling women through the reproductive lifespan (puberty, menstrual cycle, pregnancy, menopause) and endocrine-related diseases is based on studies using these techniques (Lebbe M et al, 2012). Mass spectrometry-
based assay methods have been used since 1960 for the measurement of steroid hormones, and have recently started to replace immunoassays for steroid quantification in larger clinical and research laboratories (Rosner et al., 2007). This section aims to introduce both assays, and discuss the main applications, advantages and disadvantages.
1.1.3.1 Immunoassays for steroid hormone measurements
Antibodies were first developed for large proteins, such as insulin, growth hormone, LH and FSH, which provoke immunogenic reactions when administered to experimental animals, f.ex. rabbits or mice (Yalow and Berson, 1959). Steroids are smaller molecules, and require coupling to a protein, such as albumin or thyroglobulin, to evoke antibody generation (Abraham, 1969). Steroid hormone measurement using immunoassays are subject to different issues, which might have negative effects on the sensitivity and specificity of the assay. Steroids have low affinity for the employed antibodies, and are generally present in low concentration in biological samples. Moreover, the presence of steroid molecules with similar structure induces risks of cross-reactivity (Taylor et al., 2015). The immunoassays utilizing pre- assessment extraction with organic solvent (liquid-liquid extraction) and chromatographic fractioning have shown to perform with good sensitivity and specificity (Janse et al., 2011). The more recent ‘direct’ assays have suppressed the sample purification steps, and run on automated platforms, using chemiluminescence, fluorescence or enzymatic color reactions as detection methods (Taylor et al., 2015). Although the development of monoclonal antibodies has improved the analytical performance, a lower sensitivity and specificity is often achieved. (Stanczyk, 2006). The advantages of the automated assays are their wide availability, reduced cost and ease of use. Standardization of specific assays is difficult, because of differences in antibodies and kit formulations purchased from different suppliers (Taylor et al., 2015).
1.1.3.2 Mass spectrometry-based steroid hormone measurement
The mass spectrometer is a device that measures the mass-charge ratio of charged particles (m/Q). It is composed of a sample inlet device, an ionization source, an analyzer and an ion detector. The initial steroid mixture is first separated using chromatography, either gas chromatography (GC) or liquid chromatography (LC).
GC-MS is employed to analyze the metabolites of steroid hormones and precursors. GC vaporizes liquid analytes, and separates them in a heated column based on different speed resulting from different interactions with the liquid medium in the column. Since steroid metabolites present as glucuronide and sulfate conjugates, these charged are removed by chemical derivatization, making them less polar and more volatile and stable in order to improve chromatographic resolution (McDonald et al., 2011). GC-MS is the method of reference for urinary steroid profiling, and remains at the forefront for identifying and studying steroid metabolic disorders (Krone et al., 2010)
LC-MS requires less sample preparation, and does not rely on chemical derivatization of steroid analytes. LC and High performance LC (HPLC), retain steroids in the column based on polarity of the mobile phase versus the stationary phase. The introduction of tandem mass spectrometry (MS/MS), utilizing quadrupole mass analyzers, which fragment the sample inside the instrument and analyze the generated products, have modernized the MS-field considerably (Shackleton, 2010). LC-MS/MS is the method of choice in routine clinical laboratories, because of its superior sensitivity and specificity with regard to immunoassays, the possibility to measure several different steroids simultaneously and the high throughput capacity (Grebe and Singh, 2011). The LC-MS/MS methods require thorough, standardized validation and appropriate technical training to optimize their analytical performance (Taylor et al., 2015). Importantly, the employed assay requires calibration against approved standards (Wierman et
al., 2014a). Kushnir et al. published a standardized sample preparation technique using for the quantification of steroid hormones, and made available reference ranges for testosterone in pre- and post-menopausal women (Kushnir et al., 2010).
1.1.3.3 Which method to choose?
Choosing the correct assay for research or clinical purposes might be a difficult, and will depend on the specific scientific or diagnostic question, the expected steroid concentration in the biological sample, the availability of the technique, costs and technical and analytical training (Taylor et al., 2015) A Sex Steroid Assay Reporting Task Force, issued by the Endocrine Society, has recently highlighted the requirements for analytical validity and quality of the steroid hormone assay, with regard to accuracy, sensitivity, specificity, and reproducibility (Wierman et al., 2014b). This statement arises following a vigorous debate, provoked by an editorial promoting MS assays for reporting steroid hormone measurement (Handelsman and Wartofsky, 2013) and indicates a place for both immunoassay and MS in the current state of steroid hormone measurements, although MS will eventually take the lead in the future.
For testosterone assays, the sensitivity of immunoassays is sufficient to detect circulating levels in male subjects, but inaccurate to test samples from women or children, in which case MS is recommended (Rosner et al., 2007). Another problem for androgen measurement by immunoassay is the high level of cross-reactivity between structurally similar molecules, the most abundant circulating steroid DHEAS, present in µM concentrations, will compete and interfere with low testosterone levels (<10 nM) for antibody binding (Taylor et al., 2015).
The Endocrine Society position on estradiol assays states that good quality immunoassays and validated MS methods are both convenient for measuring levels present in normal or high concentrations (such as in in vitro fertilization or IVF settings), but inaccurate for determining very low levels, as present in non-reproductive tissues or in male patients (Rosner et al., 2013). With both immunoassays and MS-based methods, current analytical performance for estradiol measuring needs improvement (Vesper et al., 2014), and in particular sensitivity in the low pM range. Both techniques have reported ultrasensitive methods, but the availability of these techniques is extremely limited (Klein et al., 1994, Owen et al., 2014).
1.1.3.4 Free testosterone levels
Sex hormone binding globulin (SHBG) and albumin are the main binding proteins for testosterone in the circulation. Free fractions of testosterone are calculated using total testosterone, concentrations of binding proteins and dissociation coefficients between testosterone and SHBG and testosterone and albumin (Sodergard et al., 1982). A commonly used calculation of free androgen index is total testosterone X 100/SHBG blood level (Vermeulen et al., 1999). The most accurate way of measuring free testosterone levels are done by equilibrium dialysis, but this technique is not routinely used in clinical practice (Vermeulen et al., 1999).