C315 Williams fig1

LC-MS/MS measurement of serum steroids in the clinical laboratory

In recent decades liquid chromatography–tandem mass spectrometry (LC-MS/MS) has become more widespread in the clinical laboratory, bridging the analytical gap between high-throughput (but interference prone) immunoassays and the highly specific (but labour intensive) technique of gas chromatography–mass spectrometry (GC-MS). This article discusses serum steroid measurement by LC-MS/MS and describes a multiplexed LC-MS/MS steroid panel recently launched at Imperial College Healthcare NHS Trust.

by Dr Emma L. Williams

Historically steroid hormones have been measured, primarily in urine, by GC-MS and in serum and plasma by radio-immunoassay. Both techniques require sample extraction prior to analysis and for the former there is a need for derivatization to form volatile derivatives. Thus the assays are laborious and time consuming and have been the preserve of research and specialist laboratories. More recently automated immunoassays have been used in routine clinical laboratories, but these are notorious for being highly prone to interference as a result of their inherent specificity problems [1]. In recent decades LC-MS/MS has come to the fore, offering a promising alternative to immunoassays for high-throughput, specific measurement of serum steroids and it is now the method of choice in many clinical laboratories. LC-MS/MS measurement of serum steroids is informative in the clinical investigation of conditions such as hirsutism, polycystic ovarian syndrome (PCOS) and infertility. In addition LC-MS/MS steroid measurement forms part of a diagnostic triad, along with urine steroid profiling by GC-MS and whole gene sequencing of genomic DNA, for inherited steroidogenic defects including the congenital adrenal hyperplasias (CAH) and disorders of sexual differentiation.

LC-MS/MS measurement
Significant advances in LC-MS/MS technology have enabled the development of high-throughput, sensitive and precise assays for steroid measurement. Figure 1 depicts the biosynthetic pathways of steroidogenesis. LC-MS/MS assays have now been published for all of the steroids in this pathway, using a variety of approaches for sample preparation prior to analysis. Protein precipitation, liquid–liquid extraction, solid phase extraction and supported liquid extraction have all been used for the preparation step. In my laboratory, semi-automated off-line solid phase extraction has been implemented in order to achieve higher throughput. This extraction approach is used to prepare samples prior to ultra-performance (UP)LC-MS/MS analysis using electrospray ionization with detection by multiple reaction monitoring (MRM). The majority of steroids are measured in positive ionization mode, although we use negative ionization mode for aldosterone and dehydroepiandrosterone sulphate (DHEAS).

For accurate LC-MS/MS quantitation, stable isotope internal standards (IS) are required. Addition of IS to all samples, calibrators and quality controls (QCs) is carried out prior to extraction and LC-MS/MS analysis. The ratios of analyte to IS signals are determined to correct for effects of the matrix upon signal intensity, which may be due to ion suppression or enhancement. Typically in LC-MS/MS assays the IS will have two or more hydrogens replaced by deuterium atoms. The IS has a different mass and ion transition to the analyte, while retaining its chemical and physical properties and thus behaves the same way as the analyte throughout the analytical procedure. Carbon-13 labelled IS are increasingly being used as they have become more available. These co-elute more completely with the non-labelled analyte and are, therefore, more effective at correcting for matrix effects compared to deuterium labelling, which alters polarity and increases the possibility of non-co-elution.

An important factor to consider in steroid LC-MS/MS assays is that of specificity, given the similarities in structures of the various steroid intermediates in the steroidogenic pathway.
There are several examples of steroids that have the same molecular weight and are, therefore, isobaric. It is vital that these isobaric steroids are chromatographically resolved as they will undergo the same ion transitions in the mass spectrometer. If not resolved, they would be measured as if they were the same steroid and, therefore, be a cause of positive interference. For example 11-deoxycortisol and 21-deoxycortisol have the same molecular weight (Fig. 2) and undergo the same ion transitions, but can be chromatographically resolved using the selectivity of the mobile phase. It can be seen in Figure 3 that these steroids are successfully resolved in our laboratory method, which uses reverse phase T3 chromatography.

LC-MS/MS steroid assays
In the clinical laboratory, testosterone is the serum steroid most frequently measured by LC-MS/MS analysis. In the external quality assessment scheme offered by the United Kingdom National External Quality Assessment Service (UK NEQAS), 43 (21%) participating labs use LC-MS/MS, with the remainder relying upon automated immunoassays. In my laboratory, both measurement techniques are used, whereby all female samples with elevated immunoassay testosterone results >2.0 nmol/L are reflexed for LC-MS/MS confirmation. In a recent audit of over 5000 female samples in which testosterone was measured we found that of over 800 elevated samples reflexed for confirmation, 23% of these are subsequently found to have normal LC-MS/MS results within the reference range. It is, therefore, essential that elevated female immunoassay results are confirmed by LC-MS/MS to avoid falsely elevated results being reported. Norethisterone, a synthetic form of progesterone used in hormonal contraceptives, is a commonly encountered cause of positive interference in immunoassays for testosterone in female samples [2].

Advantages of multiplexed assays
Testosterone is measured in the investigation of females presenting with clinical signs of hyperandrogenism, e.g. acne and hirsutism and in the investigation of infertility and PCOS. Following the introduction of LC-MS/MS assays into the clinical laboratory for the combined measurement of testosterone and androstenedione it became clear that androstenedione is the cause of hyperandrogenism in a subgroup of patients with PCOS [3]. These cases previously may have been undiagnosed when the testosterone measured in isolation was found to be normal. This observation highlights the benefits of being able to measure two or more steroids simultaneously, which is not possible with radio-immunoassays or in routine automated immunoassays.

17-Hydroxyprogesterone (17-OHP) measurement is used to screen for 21-hydroxylase deficiency; the most common cause of CAH, accounting for ~85% of cases. 17-OHP sits at a branch point for either cortisol or androgen synthesis (Fig. 1) and accumulates when 21-hydroxylase is deficient. However, it can also be raised in normal newborns, particularly in premature neonates, and is influenced by birth weight and stress. In 21-hydroxylase deficiency, 21-deoxycortisol is formed as a side product from the accumulated 17-OHP in a reaction catalysed by 11-beta hydroxylase. The LC-MS/MS measurement of 21-deoxycortisol for the diagnosis of CAH was first described by Cristoni et al. [4] and it allows accurate diagnosis of 21-hydroxylase deficiency in newborns independent of prematurity, birth weight and stress [5]. Shackleton has proposed that a second tier panel comprising 17-OHP, cortisol, 21-deoxycortisol and androstenedione is used in newborn screening for 21-hydroxylase deficiency with a third tier of urinary GC-MS analysis to clinch the final diagnosis [6]. The addition of 11-deoxycortisol to this panel permits the diagnosis of 11-beta-hydroxylase deficiency, the second most common form of CAH. Such a panel has been applied to second tier testing for CAH [7].

In my laboratory a semi-automated solid phase extraction (SPE) LC-MS/MS method for the simultaneous measurement of androstenedione, testosterone and 17-OHP has been in use since April 2016. The SPE uses Waters Oasis PRiME HLB, 96 well, μ-elution plates and is performed using a Tecan Freedom Evo automated Liquid Handler. One hundred microlitres of sample is mixed with IS and proteins are precipitated with methanol and water. Supernatants are applied to the wells of the SPE plate and drawn through under vacuum. Following washing with 0.1% formic acid in 35% methanol, steroids are eluted with methanol and water enabling direct LC-MS/MS analysis of the eluates.
Using a Waters Acquity UPLC system, samples are injected onto a Waters Acquity UPLC HSS T3 column (2.1 × 50 mm) and separated by water/methanol/ammonium acetate/formic acid gradient elution. The analysis is performed using a Waters Acquity-TQD mass spectrometer in electrospray positive ionization mode. The analytes and their co-eluting isotopic ISs are detected using MRM. Quantifier transitions (m/z) monitored are 287>97 for androstenedione, 289>97 for testosterone and 331>97 for 17-OHP.

The method underwent full validation prior to implementation according to Clinical and Laboratory Standards Institute (CLSI) guidelines and as recommended by Honour [8] and demonstrated excellent linearity over the analytical range, with all r2 values ≥0.99. Overall process efficiency was 100–108.3%, demonstrating excellent recovery and minimal ion suppression/enhancement. Intra-assay precision was 2.6–8.1% for all analytes across the measurement range, and inter-assay precision varied from 4.9 to 10.8%. Analysis of UK NEQAS samples revealed minimal negative bias and the high specificity of the assay was confirmed by spiking and interference studies. The newly developed assay compared favourably with the stand-alone LC-MS/MS methods in use previously in our laboratory, with no requirement to re-derive reference intervals. This supra-regional assay service (SAS) accredited steroid panel assay has been in routine use in our LC-MS/MS laboratory since April 2016, streamlining the analytical service. The assay is carried out two or three times a week, with each full plate accommodating around 80 patient samples, plus standards and controls, with automated sample extraction completed in ~ 90 minutes and the LC-MS/MS sample to sample injection time is 5 minutes.

We have recently evaluated a seven steroid LC-MS/MS assay with the addition of cortisol, DHEAS, 11-deoxycortisol and 21-deoxycortisol into the panel. Figure 3 shows the total ion chromatogram of the steroids quantified by this assay. Using a Waters Acquity-TQD mass spectrometer and a slightly modified experimental set-up, the lower limits of quantification obtained were 16.5 nmol/L for cortisol, 2nmol/L for DHEAS, 7nmol/L for 11-deoxycortisol and 2nmol/L for 21-deoxycortisol.
In conclusion, LC-MS/MS steroid panels are a valuable addition to the diagnostic work up of patients being investigated for hyperandrogenism and in the investigation of steroidogenic defects. The increased availability of semi-automated, high-throughput LC-MS/MS assays for multiplexed steroid measurement has opened the door for their future application in targeted metabolomic research. Finally, in the clinical laboratory setting the future continues to look bright for the role of accurate and robust measurement by LC-MS/MS in place of immunoassays as the method of choice for routine serum steroid measurement.

1. Jones AM, Honour JW. Unusual results from immunoassays and the role of the clinical endocrinologist. Clin Endocrinol Oxf 2006; 64: 234–244.
2. Jeffery J, MacKenzie F, Beckett G, Perry L, Ayling R. Norethisterone interference in testosterone assays. Ann Clin Biochem 2014; 51: 284–288.
3. Livadas S, Pappas C, Karachalios A, Marinakis E, Tolia N, Drakou M, Kaldrymides P, Panidis D, Diamanti-Kandarakis E. Prevalence and impact of hyperandrogenemia in 1218 women with polycystic ovarian syndrome. Endocrine 2014; 47: 631–638.
4. Cristoni S, Cuccato D, Sciannamblo M, Bernardi LR, Biunno I, Gerthoux P, Russo G, Weber G, Mora S. Analysis of 21-deoxycortisol, a marker of congenital adrenal hyperplasia, in blood by atmospheric pressure chemical ionization and electrospray ionization using multiple reaction monitoring. Rapid Commun Mass Spectrom 2004; 18: 77–82.
5. Janzen N, Peter M, Sander S, Steuerwald U, Terhardt M, Holtkamp U, Sander J. Newborn screening for congenital adrenal hyperplasia: additional steroid profile using liquid chromatography-tandem mass spectrometry. J Clin Endocrinol Metab 2007; 92: 2581–2589.
6. Shackleton C. Clinical steroid mass spectrometry: a 45-year history culminating in HPLC-MS/MS becoming an essential tool for patient diagnosis. J Steroid Biochem Mol Biol 2010; 121: 481–490.
7. Rossi C, Calton L, Hammond G, Brown HA, Wallace AM, Sacchetta P, Morris M. Serum steroid profiling for congenital adrenal hyperplasia using liquid chromatography-tandem mass spectrometry. Clin Chim Acta 2010; 411: 222–228.
8. Honour JW. Development and validation of a quantitative assay based on tandem mass spectrometry. Ann Clin Biochem 2011; 48: 97–111.

The author
Emma L. Williams PhD, FRCPath
North West London Pathology, Imperial College Healthcare NHS Trust, London
W6 8RF, UK

E-mail: emma.walker15@nhs.net